[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-aws-agents-get-started":3,"mdc--sgs9p6-key":32,"related-repo-aws-agents-get-started":2722,"related-org-aws-agents-get-started":2815},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":21,"repoUrl":22,"updatedAt":23,"license":24,"forks":25,"topics":26,"repo":27,"sourceUrl":30,"mdContent":31},"agents-get-started","scaffold and deploy new agent projects","Use when a developer wants to create a new agent project or get started with AgentCore. Handles framework selection, project scaffolding, first deploy, and first invocation. Triggers on: \"build an agent\", \"create an agent\", \"get started\", \"new project\", \"agentcore create\", \"which framework\", \"Strands vs LangGraph\", \"hello world agent\", \"first agent\", \"create MCP server\", \"host MCP server\", \"agentcore dev\", \"dev server\", \"what port\", \"local development\". Not for adding capabilities to existing projects — use agents-build or agents-connect. Strands vs LangGraph in a migration context routes to agents-build, not here. Connecting to an existing MCP server routes to agents-connect, not here.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"aws","AWS (Amazon)","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Faws.png",[12,16,19],{"name":13,"slug":14,"type":15},"Deployment","deployment","tag",{"name":17,"slug":18,"type":15},"Agents","agents",{"name":20,"slug":8,"type":15},"AWS",1822,"https:\u002F\u002Fgithub.com\u002Faws\u002Fagent-toolkit-for-aws","2026-07-12T08:42:51.963247",null,157,[],{"repoUrl":22,"stars":21,"forks":25,"topics":28,"description":29},[],"Official, AWS-supported MCP servers, skills, and plugins to help AI agents build on AWS","https:\u002F\u002Fgithub.com\u002Faws\u002Fagent-toolkit-for-aws\u002Ftree\u002FHEAD\u002Fplugins\u002Faws-agents\u002Fskills\u002Fagents-get-started","---\nname: agents-get-started\ndescription: >\n  Use when a developer wants to create a new agent project or get started\n  with AgentCore. Handles framework selection, project scaffolding, first\n  deploy, and first invocation. Triggers on: \"build an agent\", \"create an\n  agent\", \"get started\", \"new project\", \"agentcore create\", \"which\n  framework\", \"Strands vs LangGraph\", \"hello world agent\", \"first agent\",\n  \"create MCP server\", \"host MCP server\", \"agentcore dev\", \"dev server\",\n  \"what port\", \"local development\".\n  Not for adding capabilities to existing projects — use agents-build\n  or agents-connect. Strands vs LangGraph in a migration context routes\n  to agents-build, not here. Connecting to an existing MCP server routes\n  to agents-connect, not here.\nallowed-tools: Read Grep Glob Bash\nmetadata:\n  type: skill\n  version: \"1.0.0\"\n  author: aws-agentcore\n  requires-cli: \">=0.9.0\"\n---\n\n# get-started\n\nWalk a developer from zero to a running agent on AWS.\n\n## When to use\n\n- Developer wants to build an agent on AWS and doesn't know where to start\n- Developer wants to create a new AgentCore project\n- Developer is choosing between frameworks (Strands, LangGraph, GoogleADK, OpenAI Agents)\n- Developer just ran `agentcore create` and wants to know what to do next\n\nDo NOT use for:\n\n- Environment\u002Fprerequisite issues (CLI not found, credentials broken) → use `agents-debug`\n- Adding capabilities to an existing project (memory, tools, policies) → use `agents-build` or `agents-connect`\n- Migrating an existing Bedrock Agent → use `agents-build` (loads [`references\u002Fmigrate.md`](..\u002Fagents-build\u002Freferences\u002Fmigrate.md))\n\n## Input\n\n`$ARGUMENTS` can be:\n\n- A framework preference: \"using LangGraph\", \"with Strands\"\n- A protocol: \"MCP server\", \"A2A\"\n- A description of what the agent should do: \"a customer support agent\"\n- Empty — the skill will guide framework selection\n\n## Process\n\n### Step 0: Verify CLI version\n\n```bash\nagentcore --version\n```\n\nThis skill requires v0.9.0 or later.\n\nIf the version is older:\n> Your AgentCore CLI is out of date (found vX.Y.Z, need v0.9.0+).\n\nOffer to run the update: `agentcore update`. After the update completes, re-check the version to confirm it's ≥0.9.0 before continuing. Preserve any context the developer already provided (framework preference, project name, what they want to build) so they don't have to repeat themselves.\n\nIf `agentcore` is not found:\n> The AgentCore CLI isn't installed. Run `npm install -g @aws\u002Fagentcore` (requires Node.js 20+).\n> If you're having trouble with installation, I can run the `agents-debug` skill (which loads [`references\u002Fdoctor.md`](..\u002Fagents-debug\u002Freferences\u002Fdoctor.md)) to diagnose your environment.\n\n### Step 1: Determine intent — exploring or ready to create?\n\nBefore jumping into framework selection, figure out where the developer is:\n\n**Ask the developer:** \"Are you exploring options (comparing frameworks, understanding what AgentCore does) or ready to create a project?\"\n\n- **Exploring** → Go to Step 2 (framework comparison). Present the options, answer questions, and wait. Do not construct a `create` command until they signal they're ready.\n- **Ready to create** → Skip to Step 3 (create the project). If they already specified a framework, skip Step 2 entirely.\n- **Already has a project** → Look for `agentcore\u002Fagentcore.json` in the current directory. If found, read it and skip to Step 5 (what to do next). Don't re-scaffold.\n\nIf the developer's intent is clear from `$ARGUMENTS` (e.g., \"create a Strands agent called MyBot\"), skip straight to Step 3.\n\n### Step 2: Framework selection\n\n**Check conversation context first.** If the developer already discussed frameworks earlier in this conversation (e.g., from a previous skill invocation), don't re-present the full table. Summarize what was discussed and ask if they've decided, or if anything changed.\n\nIf this is the first time discussing frameworks, present the options:\n\n**Supported frameworks (CLI-scaffolded, Python):**\n\n| Framework | CLI value | Best for |\n|---|---|---|\n| Strands | `Strands` | AWS-native, simplest path, best AgentCore integration |\n| LangGraph | `LangChain_LangGraph` | Complex graph-based workflows, existing LangChain investment |\n| Google ADK | `GoogleADK` | Teams already using Google's agent toolkit |\n| OpenAI Agents | `OpenAIAgents` | Teams already using OpenAI's agent SDK |\n\n**Ask the developer to choose.** Present the options and wait for their selection. Don't assume a default unless they explicitly say they have no preference.\n\n> **Note on naming:** The CLI flag value is the exact string to pass to `--framework`. In prose use the shorter names.\n\n**Default recommendation** (only when the developer says \"no preference\" or \"you pick\"): Strands — AWS-native framework with the tightest AgentCore integration and the most samples\u002Fdocs.\n\n**Key decision points to surface:**\n\n- \"Do you have existing agent code in LangGraph or OpenAI Agents?\" → use that framework\n- \"Do you need complex graph-based workflows with conditional branching?\" → LangGraph\n- \"Starting fresh with no preference?\" → Strands\n\n#### Framework not listed?\n\nIf the developer asks about a framework not in the table above, handle it:\n\n| They ask about | What to say |\n|---|---|\n| **CrewAI, AutoGen, Semantic Kernel** | Not scaffolded by the CLI, but you can use them via the BYO Container path (below). AgentCore Runtime is framework-agnostic — any code that implements the HTTP contract works. |\n| **Anthropic SDK \u002F Claude Agent SDK** | This is a model SDK, not an agent framework. You can use it inside any framework (Strands, LangGraph, etc.) or standalone. For standalone use, wrap it in a container with the Runtime contract. |\n| **Claude Code \u002F Cursor \u002F Copilot** | These are IDE tools, not agent frameworks. They're where you *write* agent code, not what you deploy. Pick a framework from the table above for the agent itself. |\n| **LangChain (without LangGraph)** | LangChain is a library, LangGraph is the agent framework built on it. The CLI scaffolds LangGraph. If you're using plain LangChain chains, the BYO Container path works. |\n| **Custom \u002F homegrown framework** | BYO Container path — see below. |\n\n**BYO Container path (any framework, any language):**\n\nFor frameworks or languages not scaffolded by the CLI, AgentCore Runtime accepts any container that implements the HTTP contract (`POST \u002Finvocations`, `GET \u002Fping`). The workflow:\n\n1. `agentcore create --name \u003CProjectName> --defaults` to scaffold the project structure\n2. `agentcore add agent --type byo --build Container --language \u003CLanguage> --code-location \u003Cpath>` to register your code\n3. Write a `Dockerfile` that builds and runs your agent\n4. `agentcore deploy` handles ECR push, CDK infra, and runtime creation\n\n**Language-specific notes:**\n\n| Language | Recommended path |\n|---|---|\n| Java (Spring Boot) | [Spring AI SDK for AgentCore](https:\u002F\u002Faws.amazon.com\u002Fblogs\u002Fmachine-learning\u002Fspring-ai-sdk-for-amazon-bedrock-agentcore-is-now-generally-available) — handles the Runtime contract, SSE streaming, and health checks. Use `--language Other --build Container`. |\n| JavaScript \u002F TypeScript | Implement the Runtime contract in Express\u002FFastify\u002Fetc. Use `--language TypeScript --build Container`. |\n| Go, Rust, .NET, other | Implement the Runtime HTTP contract. Use `--language Other --build Container`. |\n\nThe rest of this skill (deploy, status, logs, invoke) applies once the container builds correctly.\n\n#### Framework vs. model provider — a common confusion\n\nThe framework is how your agent orchestrates (Strands, LangGraph, etc.). The model provider is which LLM it calls (Bedrock, Anthropic, OpenAI, Gemini). These are independent choices:\n\n- Strands + Bedrock (default) — AWS-native everything\n- Strands + Anthropic — Strands orchestration, direct Anthropic API for the model\n- LangGraph + Bedrock — LangGraph orchestration, Bedrock for the model\n- OpenAI Agents + OpenAI — OpenAI everything\n\nIf the developer says \"I want to use Claude\" they mean the model provider (Bedrock or Anthropic), not the framework. If they say \"I want to use LangGraph\" they mean the framework.\n\n### Step 3: Create the project\n\nBuild the `agentcore create` command based on the developer's choices.\n\n**Before constructing the command — validate the project name.** The CLI fails late: if the name is invalid, you'll see the error *after* walking through prompts or building the full command. Save the round-trip and check these rules up front. Reject the name and ask for a new one if any rule fails:\n\n- **Length ≤ 23 characters** (this is shorter than most developers assume — `MyCustomerSupportAgent` is 22 chars and fits; `CustomerSupportChatbot` is 22 and fits; `MyCustomerSupportBotApp` is 23 and just fits; `MyCustomerSupportChatBot` is 24 and **fails**)\n- **Alphanumeric only** — no hyphens, underscores, dots, or spaces\n- **Must start with a letter**\n\nSay the count back out loud when close to the limit: \"That name is 24 characters — the CLI caps project names at 23. Want to shorten it to `\u003Csuggestion>`?\" Do not run the command with an invalid name on the assumption that the CLI error message will be clear — it isn't always, and the developer's mental model will be wrong for subsequent commands.\n\n**Construct the command, then present it for confirmation before the developer runs it.** Show the full command with all flags and explain what each choice means. Wait for the developer to confirm or adjust before proceeding.\n\nExample presentation:\n\n> Here's the command I'd recommend based on what you've told me:\n>\n> ```bash\n> agentcore create --name MyAgent --framework Strands --model-provider Bedrock --build CodeZip --memory none\n> ```\n>\n> This creates a Strands agent using Bedrock models, deployed as a code zip (no Docker needed). Memory can be added later.\n>\n> Want to run this, or change anything?\n\nDo NOT execute the command automatically — present it and wait.\n\n**Minimal (defaults — Strands, Bedrock, CodeZip, no memory):**\n\n```bash\nagentcore create --name \u003CProjectName> --defaults\n```\n\n**With specific options:**\n\n```bash\nagentcore create \\\n  --name \u003CProjectName> \\\n  --framework \u003CFramework> \\\n  --model-provider Bedrock \\\n  --build CodeZip \\\n  --memory none\n```\n\n**Flag reference:**\n\n| Flag | Values | Default |\n|---|---|---|\n| `--name` | alphanumeric, max 23 chars | prompted |\n| `--framework` | `Strands`, `LangChain_LangGraph`, `GoogleADK`, `OpenAIAgents` | prompted |\n| `--protocol` | `HTTP`, `MCP`, `A2A` | `HTTP` |\n| `--build` | `CodeZip`, `Container` | `CodeZip` |\n| `--model-provider` | `Bedrock`, `Anthropic`, `OpenAI`, `Gemini` | prompted |\n| `--memory` | `none`, `shortTerm`, `longAndShortTerm` | prompted |\n| `--network-mode` | `PUBLIC`, `VPC` | `PUBLIC` |\n| `--dry-run` | — | preview without creating |\n\n**Guidance on choices:**\n\n- **Protocol:** Use `HTTP` unless the developer specifically needs MCP tool serving or A2A agent-to-agent communication\n- **Build:** Use `CodeZip` unless the developer needs custom system dependencies (CodeZip is faster to deploy and doesn't require Docker locally)\n- **Model provider:** Use `Bedrock` unless the developer has a specific reason for another provider (Bedrock doesn't require managing API keys)\n- **Memory:** Start with `none` — memory can be added later via `agents-build` (loads [`references\u002Fmemory.md`](..\u002Fagents-build\u002Freferences\u002Fmemory.md)) when the developer needs it\n\n### Step 4: Explain what was created\n\nAfter the project exists, read `agentcore\u002Fagentcore.json` and the generated code to explain the project structure.\n\nThe layout below reflects CLI v0.9.x. If the CLI version is different, run `tree \u003CProjectName>\u002F -L 3` to see the actual generated structure and explain from there.\n\n```\n\u003CProjectName>\u002F\n├── agentcore\u002F\n│   ├── agentcore.json      ← Project config (agents, resources)\n│   ├── aws-targets.json    ← AWS account + region\n│   ├── .env.local          ← Local environment variables (gitignored)\n│   └── cdk\u002F                ← CDK infrastructure (auto-managed, don't edit)\n└── app\u002F\n    └── \u003CAgentName>\u002F\n        ├── main.py          ← Your agent code — this is where you build\n        ├── mcp_client\u002F      ← Pre-wired example MCP client (see note below)\n        └── pyproject.toml   ← Python dependencies\n```\n\n**Key files to highlight:**\n\n- `app\u002F\u003CAgentName>\u002Fmain.py` — the agent's entry point. This is where the developer adds tools, system prompts, and logic.\n- `agentcore\u002Fagentcore.json` — the project config. Resources are added here via `agentcore add` commands.\n- `agentcore\u002F.env.local` — local environment variables. After deploy, resource IDs are written here for local dev.\n\n**Heads-up on the scaffolded MCP client.** `main.py` imports `get_streamable_http_mcp_client()` from `mcp_client\u002Fclient.py` and appends it to `tools`. In a fresh project, this client points at a public example MCP endpoint — so `agentcore dev` works immediately. Two things to flag:\n\n1. **It will become a silent no-op if you repoint it at a gateway that isn't deployed yet.** The common path is to swap the example endpoint for `os.getenv(\"AGENTCORE_GATEWAY_\u003CNAME>_URL\")`. That env var is only populated after `agentcore deploy`. If the developer repoints and runs `agentcore dev` before deploying, `get_streamable_http_mcp_client()` returns a client with a `None` URL and the agent starts with zero MCP tools — no error, no warning. See the \"Local dev gap\" section in `agents-connect` for the guard pattern: `if not GATEWAY_URL: tools = []`.\n2. **If the developer doesn't need MCP tools at all**, remove the `mcp_clients` list and the loop that appends it to `tools`. The scaffold includes it as a convenience, not a requirement.\n\nThe reference client code in `agents-connect` (Path A) shows the correct pattern for gateway-backed MCP clients once deploy has run.\n\n### Step 5: Local development\n\n```bash\nagentcore dev\n```\n\nThis starts a local dev server. The developer can interact with their agent immediately.\n\n**Port the dev server binds to** (important if you're scripting `curl` calls or testing from another process):\n\n| Protocol | Default port |\n|---|---|\n| HTTP | `8080` |\n| MCP | `8000` |\n| A2A | `9000` |\n\nThe CLI prints the bound port and URL on startup — always read the actual value from the CLI output rather than hardcoding. **If the default port is already in use**, the CLI auto-increments (e.g., 8080 → 8081 → 8082), so a second dev session or a lingering process from a previous run can shift your port without warning. Use `agentcore dev --port \u003CN>` to pin it, or grep `ps` \u002F check the CLI banner if invocations start failing with connection-refused or exit-code-7 errors.\n\n**Important limitations to mention:**\n\n- Memory is not available in `agentcore dev` — it requires a deploy\n- Gateway URLs are not available locally — they require a deploy\n- The local server uses the model provider configured in the project\n\n### Step 6: First deploy\n\nWhen the developer is ready to deploy:\n\n```bash\nagentcore deploy\n```\n\nThis will:\n\n1. Show a preview of AWS resources to be created\n2. Ask for confirmation\n3. Build and deploy via CDK\n\n**First deploy takes 3-5 minutes.** Subsequent deploys are faster.\n\nAfter deploy, show them how to invoke:\n\n```bash\nagentcore invoke \"Hello, what can you do?\"\n```\n\nAnd how to check status:\n\n```bash\nagentcore status\n```\n\n### Step 7: What's next\n\nBased on what the developer said they want to build, suggest the logical next skill:\n\n| Developer intent | Next skill | Command hint |\n|---|---|---|\n| \"How do I call it from my app?\" | `agents-build` | `agentcore fetch access` |\n| \"I want it to remember things\" | `agents-build` | `agentcore add memory` |\n| \"I want it to call external APIs\" | `agents-connect` | `agentcore add gateway` |\n| \"I want to restrict what it can do\" | `agents-connect` | `agentcore add policy-engine` |\n| \"I want to measure quality\" | `agents-optimize` | `agentcore add evaluator` |\n| \"I want to go to production\" | `agents-harden` | production readiness checklist |\n| \"I want multiple agents working together\" | `agents-build` | `agentcore create --protocol A2A` |\n| \"I need it in a VPC\" | `agents-build` | `agentcore create --network-mode VPC` |\n\nDon't overwhelm — suggest one or two next steps based on what the developer actually asked for.\n\n### Example walkthroughs\n\nFor task-framed prompts (e.g., \"build a customer support agent\"), load the matching example reference:\n\n| Developer task | Reference |\n|---|---|\n| Customer support, chatbot, answer policy questions | [`references\u002Fexample-support-agent.md`](references\u002Fexample-support-agent.md) |\n\nMore examples can be added to this skill's references directory as common patterns emerge.\n\n## Output\n\n- A clear path from \"I want to build an agent\" to a running deployed agent\n- The `agentcore create` command tailored to their choices\n- An explanation of the generated project structure\n- Concrete next steps based on their intent\n\n## Quality criteria\n\n- The `agentcore create` command uses only valid flags from CLI v0.9.1\n- Framework recommendation is based on the developer's context, not a generic default\n- The developer understands what each generated file does\n- Next steps are specific to what the developer wants to build, not a generic list of all features\n",{"data":33,"body":40},{"name":4,"description":6,"allowed-tools":34,"metadata":35},"Read Grep Glob Bash",{"type":36,"version":37,"author":38,"requires-cli":39},"skill","1.0.0","aws-agentcore",">=0.9.0",{"type":41,"children":42},"root",[43,51,57,64,98,103,161,167,178,201,207,214,243,248,253,262,275,287,322,328,333,344,393,405,411,421,426,434,554,564,585,595,603,621,628,633,743,751,772,822,830,920,925,931,936,959,964,970,982,999,1068,1081,1091,1096,1183,1188,1196,1245,1253,1383,1391,1703,1711,1798,1804,1816,1829,1839,1847,1890,1940,2029,2041,2047,2066,2071,2089,2159,2187,2195,2220,2226,2231,2250,2255,2273,2283,2288,2322,2327,2346,2352,2357,2582,2587,2593,2598,2640,2645,2651,2681,2687,2716],{"type":44,"tag":45,"props":46,"children":48},"element","h1",{"id":47},"get-started",[49],{"type":50,"value":47},"text",{"type":44,"tag":52,"props":53,"children":54},"p",{},[55],{"type":50,"value":56},"Walk a developer from zero to a running agent on AWS.",{"type":44,"tag":58,"props":59,"children":61},"h2",{"id":60},"when-to-use",[62],{"type":50,"value":63},"When to use",{"type":44,"tag":65,"props":66,"children":67},"ul",{},[68,74,79,84],{"type":44,"tag":69,"props":70,"children":71},"li",{},[72],{"type":50,"value":73},"Developer wants to build an agent on AWS and doesn't know where to start",{"type":44,"tag":69,"props":75,"children":76},{},[77],{"type":50,"value":78},"Developer wants to create a new AgentCore project",{"type":44,"tag":69,"props":80,"children":81},{},[82],{"type":50,"value":83},"Developer is choosing between frameworks (Strands, LangGraph, GoogleADK, OpenAI Agents)",{"type":44,"tag":69,"props":85,"children":86},{},[87,89,96],{"type":50,"value":88},"Developer just ran ",{"type":44,"tag":90,"props":91,"children":93},"code",{"className":92},[],[94],{"type":50,"value":95},"agentcore create",{"type":50,"value":97}," and wants to know what to do next",{"type":44,"tag":52,"props":99,"children":100},{},[101],{"type":50,"value":102},"Do NOT use for:",{"type":44,"tag":65,"props":104,"children":105},{},[106,117,136],{"type":44,"tag":69,"props":107,"children":108},{},[109,111],{"type":50,"value":110},"Environment\u002Fprerequisite issues (CLI not found, credentials broken) → use ",{"type":44,"tag":90,"props":112,"children":114},{"className":113},[],[115],{"type":50,"value":116},"agents-debug",{"type":44,"tag":69,"props":118,"children":119},{},[120,122,128,130],{"type":50,"value":121},"Adding capabilities to an existing project (memory, tools, policies) → use ",{"type":44,"tag":90,"props":123,"children":125},{"className":124},[],[126],{"type":50,"value":127},"agents-build",{"type":50,"value":129}," or ",{"type":44,"tag":90,"props":131,"children":133},{"className":132},[],[134],{"type":50,"value":135},"agents-connect",{"type":44,"tag":69,"props":137,"children":138},{},[139,141,146,148,159],{"type":50,"value":140},"Migrating an existing Bedrock Agent → use ",{"type":44,"tag":90,"props":142,"children":144},{"className":143},[],[145],{"type":50,"value":127},{"type":50,"value":147}," (loads ",{"type":44,"tag":149,"props":150,"children":152},"a",{"href":151},"..\u002Fagents-build\u002Freferences\u002Fmigrate.md",[153],{"type":44,"tag":90,"props":154,"children":156},{"className":155},[],[157],{"type":50,"value":158},"references\u002Fmigrate.md",{"type":50,"value":160},")",{"type":44,"tag":58,"props":162,"children":164},{"id":163},"input",[165],{"type":50,"value":166},"Input",{"type":44,"tag":52,"props":168,"children":169},{},[170,176],{"type":44,"tag":90,"props":171,"children":173},{"className":172},[],[174],{"type":50,"value":175},"$ARGUMENTS",{"type":50,"value":177}," can be:",{"type":44,"tag":65,"props":179,"children":180},{},[181,186,191,196],{"type":44,"tag":69,"props":182,"children":183},{},[184],{"type":50,"value":185},"A framework preference: \"using LangGraph\", \"with Strands\"",{"type":44,"tag":69,"props":187,"children":188},{},[189],{"type":50,"value":190},"A protocol: \"MCP server\", \"A2A\"",{"type":44,"tag":69,"props":192,"children":193},{},[194],{"type":50,"value":195},"A description of what the agent should do: \"a customer support agent\"",{"type":44,"tag":69,"props":197,"children":198},{},[199],{"type":50,"value":200},"Empty — the skill will guide framework selection",{"type":44,"tag":58,"props":202,"children":204},{"id":203},"process",[205],{"type":50,"value":206},"Process",{"type":44,"tag":208,"props":209,"children":211},"h3",{"id":210},"step-0-verify-cli-version",[212],{"type":50,"value":213},"Step 0: Verify CLI version",{"type":44,"tag":215,"props":216,"children":221},"pre",{"className":217,"code":218,"language":219,"meta":220,"style":220},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","agentcore --version\n","bash","",[222],{"type":44,"tag":90,"props":223,"children":224},{"__ignoreMap":220},[225],{"type":44,"tag":226,"props":227,"children":230},"span",{"class":228,"line":229},"line",1,[231,237],{"type":44,"tag":226,"props":232,"children":234},{"style":233},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[235],{"type":50,"value":236},"agentcore",{"type":44,"tag":226,"props":238,"children":240},{"style":239},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[241],{"type":50,"value":242}," --version\n",{"type":44,"tag":52,"props":244,"children":245},{},[246],{"type":50,"value":247},"This skill requires v0.9.0 or later.",{"type":44,"tag":52,"props":249,"children":250},{},[251],{"type":50,"value":252},"If the version is older:",{"type":44,"tag":254,"props":255,"children":256},"blockquote",{},[257],{"type":44,"tag":52,"props":258,"children":259},{},[260],{"type":50,"value":261},"Your AgentCore CLI is out of date (found vX.Y.Z, need v0.9.0+).",{"type":44,"tag":52,"props":263,"children":264},{},[265,267,273],{"type":50,"value":266},"Offer to run the update: ",{"type":44,"tag":90,"props":268,"children":270},{"className":269},[],[271],{"type":50,"value":272},"agentcore update",{"type":50,"value":274},". After the update completes, re-check the version to confirm it's ≥0.9.0 before continuing. Preserve any context the developer already provided (framework preference, project name, what they want to build) so they don't have to repeat themselves.",{"type":44,"tag":52,"props":276,"children":277},{},[278,280,285],{"type":50,"value":279},"If ",{"type":44,"tag":90,"props":281,"children":283},{"className":282},[],[284],{"type":50,"value":236},{"type":50,"value":286}," is not found:",{"type":44,"tag":254,"props":288,"children":289},{},[290],{"type":44,"tag":52,"props":291,"children":292},{},[293,295,301,303,308,310,320],{"type":50,"value":294},"The AgentCore CLI isn't installed. Run ",{"type":44,"tag":90,"props":296,"children":298},{"className":297},[],[299],{"type":50,"value":300},"npm install -g @aws\u002Fagentcore",{"type":50,"value":302}," (requires Node.js 20+).\nIf you're having trouble with installation, I can run the ",{"type":44,"tag":90,"props":304,"children":306},{"className":305},[],[307],{"type":50,"value":116},{"type":50,"value":309}," skill (which loads ",{"type":44,"tag":149,"props":311,"children":313},{"href":312},"..\u002Fagents-debug\u002Freferences\u002Fdoctor.md",[314],{"type":44,"tag":90,"props":315,"children":317},{"className":316},[],[318],{"type":50,"value":319},"references\u002Fdoctor.md",{"type":50,"value":321},") to diagnose your environment.",{"type":44,"tag":208,"props":323,"children":325},{"id":324},"step-1-determine-intent-exploring-or-ready-to-create",[326],{"type":50,"value":327},"Step 1: Determine intent — exploring or ready to create?",{"type":44,"tag":52,"props":329,"children":330},{},[331],{"type":50,"value":332},"Before jumping into framework selection, figure out where the developer is:",{"type":44,"tag":52,"props":334,"children":335},{},[336,342],{"type":44,"tag":337,"props":338,"children":339},"strong",{},[340],{"type":50,"value":341},"Ask the developer:",{"type":50,"value":343}," \"Are you exploring options (comparing frameworks, understanding what AgentCore does) or ready to create a project?\"",{"type":44,"tag":65,"props":345,"children":346},{},[347,365,375],{"type":44,"tag":69,"props":348,"children":349},{},[350,355,357,363],{"type":44,"tag":337,"props":351,"children":352},{},[353],{"type":50,"value":354},"Exploring",{"type":50,"value":356}," → Go to Step 2 (framework comparison). Present the options, answer questions, and wait. Do not construct a ",{"type":44,"tag":90,"props":358,"children":360},{"className":359},[],[361],{"type":50,"value":362},"create",{"type":50,"value":364}," command until they signal they're ready.",{"type":44,"tag":69,"props":366,"children":367},{},[368,373],{"type":44,"tag":337,"props":369,"children":370},{},[371],{"type":50,"value":372},"Ready to create",{"type":50,"value":374}," → Skip to Step 3 (create the project). If they already specified a framework, skip Step 2 entirely.",{"type":44,"tag":69,"props":376,"children":377},{},[378,383,385,391],{"type":44,"tag":337,"props":379,"children":380},{},[381],{"type":50,"value":382},"Already has a project",{"type":50,"value":384}," → Look for ",{"type":44,"tag":90,"props":386,"children":388},{"className":387},[],[389],{"type":50,"value":390},"agentcore\u002Fagentcore.json",{"type":50,"value":392}," in the current directory. If found, read it and skip to Step 5 (what to do next). Don't re-scaffold.",{"type":44,"tag":52,"props":394,"children":395},{},[396,398,403],{"type":50,"value":397},"If the developer's intent is clear from ",{"type":44,"tag":90,"props":399,"children":401},{"className":400},[],[402],{"type":50,"value":175},{"type":50,"value":404}," (e.g., \"create a Strands agent called MyBot\"), skip straight to Step 3.",{"type":44,"tag":208,"props":406,"children":408},{"id":407},"step-2-framework-selection",[409],{"type":50,"value":410},"Step 2: Framework selection",{"type":44,"tag":52,"props":412,"children":413},{},[414,419],{"type":44,"tag":337,"props":415,"children":416},{},[417],{"type":50,"value":418},"Check conversation context first.",{"type":50,"value":420}," If the developer already discussed frameworks earlier in this conversation (e.g., from a previous skill invocation), don't re-present the full table. Summarize what was discussed and ask if they've decided, or if anything changed.",{"type":44,"tag":52,"props":422,"children":423},{},[424],{"type":50,"value":425},"If this is the first time discussing frameworks, present the options:",{"type":44,"tag":52,"props":427,"children":428},{},[429],{"type":44,"tag":337,"props":430,"children":431},{},[432],{"type":50,"value":433},"Supported frameworks (CLI-scaffolded, Python):",{"type":44,"tag":435,"props":436,"children":437},"table",{},[438,462],{"type":44,"tag":439,"props":440,"children":441},"thead",{},[442],{"type":44,"tag":443,"props":444,"children":445},"tr",{},[446,452,457],{"type":44,"tag":447,"props":448,"children":449},"th",{},[450],{"type":50,"value":451},"Framework",{"type":44,"tag":447,"props":453,"children":454},{},[455],{"type":50,"value":456},"CLI value",{"type":44,"tag":447,"props":458,"children":459},{},[460],{"type":50,"value":461},"Best for",{"type":44,"tag":463,"props":464,"children":465},"tbody",{},[466,488,510,532],{"type":44,"tag":443,"props":467,"children":468},{},[469,475,483],{"type":44,"tag":470,"props":471,"children":472},"td",{},[473],{"type":50,"value":474},"Strands",{"type":44,"tag":470,"props":476,"children":477},{},[478],{"type":44,"tag":90,"props":479,"children":481},{"className":480},[],[482],{"type":50,"value":474},{"type":44,"tag":470,"props":484,"children":485},{},[486],{"type":50,"value":487},"AWS-native, simplest path, best AgentCore integration",{"type":44,"tag":443,"props":489,"children":490},{},[491,496,505],{"type":44,"tag":470,"props":492,"children":493},{},[494],{"type":50,"value":495},"LangGraph",{"type":44,"tag":470,"props":497,"children":498},{},[499],{"type":44,"tag":90,"props":500,"children":502},{"className":501},[],[503],{"type":50,"value":504},"LangChain_LangGraph",{"type":44,"tag":470,"props":506,"children":507},{},[508],{"type":50,"value":509},"Complex graph-based workflows, existing LangChain investment",{"type":44,"tag":443,"props":511,"children":512},{},[513,518,527],{"type":44,"tag":470,"props":514,"children":515},{},[516],{"type":50,"value":517},"Google ADK",{"type":44,"tag":470,"props":519,"children":520},{},[521],{"type":44,"tag":90,"props":522,"children":524},{"className":523},[],[525],{"type":50,"value":526},"GoogleADK",{"type":44,"tag":470,"props":528,"children":529},{},[530],{"type":50,"value":531},"Teams already using Google's agent toolkit",{"type":44,"tag":443,"props":533,"children":534},{},[535,540,549],{"type":44,"tag":470,"props":536,"children":537},{},[538],{"type":50,"value":539},"OpenAI Agents",{"type":44,"tag":470,"props":541,"children":542},{},[543],{"type":44,"tag":90,"props":544,"children":546},{"className":545},[],[547],{"type":50,"value":548},"OpenAIAgents",{"type":44,"tag":470,"props":550,"children":551},{},[552],{"type":50,"value":553},"Teams already using OpenAI's agent SDK",{"type":44,"tag":52,"props":555,"children":556},{},[557,562],{"type":44,"tag":337,"props":558,"children":559},{},[560],{"type":50,"value":561},"Ask the developer to choose.",{"type":50,"value":563}," Present the options and wait for their selection. Don't assume a default unless they explicitly say they have no preference.",{"type":44,"tag":254,"props":565,"children":566},{},[567],{"type":44,"tag":52,"props":568,"children":569},{},[570,575,577,583],{"type":44,"tag":337,"props":571,"children":572},{},[573],{"type":50,"value":574},"Note on naming:",{"type":50,"value":576}," The CLI flag value is the exact string to pass to ",{"type":44,"tag":90,"props":578,"children":580},{"className":579},[],[581],{"type":50,"value":582},"--framework",{"type":50,"value":584},". In prose use the shorter names.",{"type":44,"tag":52,"props":586,"children":587},{},[588,593],{"type":44,"tag":337,"props":589,"children":590},{},[591],{"type":50,"value":592},"Default recommendation",{"type":50,"value":594}," (only when the developer says \"no preference\" or \"you pick\"): Strands — AWS-native framework with the tightest AgentCore integration and the most samples\u002Fdocs.",{"type":44,"tag":52,"props":596,"children":597},{},[598],{"type":44,"tag":337,"props":599,"children":600},{},[601],{"type":50,"value":602},"Key decision points to surface:",{"type":44,"tag":65,"props":604,"children":605},{},[606,611,616],{"type":44,"tag":69,"props":607,"children":608},{},[609],{"type":50,"value":610},"\"Do you have existing agent code in LangGraph or OpenAI Agents?\" → use that framework",{"type":44,"tag":69,"props":612,"children":613},{},[614],{"type":50,"value":615},"\"Do you need complex graph-based workflows with conditional branching?\" → LangGraph",{"type":44,"tag":69,"props":617,"children":618},{},[619],{"type":50,"value":620},"\"Starting fresh with no preference?\" → Strands",{"type":44,"tag":622,"props":623,"children":625},"h4",{"id":624},"framework-not-listed",[626],{"type":50,"value":627},"Framework not listed?",{"type":44,"tag":52,"props":629,"children":630},{},[631],{"type":50,"value":632},"If the developer asks about a framework not in the table above, handle it:",{"type":44,"tag":435,"props":634,"children":635},{},[636,652],{"type":44,"tag":439,"props":637,"children":638},{},[639],{"type":44,"tag":443,"props":640,"children":641},{},[642,647],{"type":44,"tag":447,"props":643,"children":644},{},[645],{"type":50,"value":646},"They ask about",{"type":44,"tag":447,"props":648,"children":649},{},[650],{"type":50,"value":651},"What to say",{"type":44,"tag":463,"props":653,"children":654},{},[655,671,687,711,727],{"type":44,"tag":443,"props":656,"children":657},{},[658,666],{"type":44,"tag":470,"props":659,"children":660},{},[661],{"type":44,"tag":337,"props":662,"children":663},{},[664],{"type":50,"value":665},"CrewAI, AutoGen, Semantic Kernel",{"type":44,"tag":470,"props":667,"children":668},{},[669],{"type":50,"value":670},"Not scaffolded by the CLI, but you can use them via the BYO Container path (below). AgentCore Runtime is framework-agnostic — any code that implements the HTTP contract works.",{"type":44,"tag":443,"props":672,"children":673},{},[674,682],{"type":44,"tag":470,"props":675,"children":676},{},[677],{"type":44,"tag":337,"props":678,"children":679},{},[680],{"type":50,"value":681},"Anthropic SDK \u002F Claude Agent SDK",{"type":44,"tag":470,"props":683,"children":684},{},[685],{"type":50,"value":686},"This is a model SDK, not an agent framework. You can use it inside any framework (Strands, LangGraph, etc.) or standalone. For standalone use, wrap it in a container with the Runtime contract.",{"type":44,"tag":443,"props":688,"children":689},{},[690,698],{"type":44,"tag":470,"props":691,"children":692},{},[693],{"type":44,"tag":337,"props":694,"children":695},{},[696],{"type":50,"value":697},"Claude Code \u002F Cursor \u002F Copilot",{"type":44,"tag":470,"props":699,"children":700},{},[701,703,709],{"type":50,"value":702},"These are IDE tools, not agent frameworks. They're where you ",{"type":44,"tag":704,"props":705,"children":706},"em",{},[707],{"type":50,"value":708},"write",{"type":50,"value":710}," agent code, not what you deploy. Pick a framework from the table above for the agent itself.",{"type":44,"tag":443,"props":712,"children":713},{},[714,722],{"type":44,"tag":470,"props":715,"children":716},{},[717],{"type":44,"tag":337,"props":718,"children":719},{},[720],{"type":50,"value":721},"LangChain (without LangGraph)",{"type":44,"tag":470,"props":723,"children":724},{},[725],{"type":50,"value":726},"LangChain is a library, LangGraph is the agent framework built on it. The CLI scaffolds LangGraph. If you're using plain LangChain chains, the BYO Container path works.",{"type":44,"tag":443,"props":728,"children":729},{},[730,738],{"type":44,"tag":470,"props":731,"children":732},{},[733],{"type":44,"tag":337,"props":734,"children":735},{},[736],{"type":50,"value":737},"Custom \u002F homegrown framework",{"type":44,"tag":470,"props":739,"children":740},{},[741],{"type":50,"value":742},"BYO Container path — see below.",{"type":44,"tag":52,"props":744,"children":745},{},[746],{"type":44,"tag":337,"props":747,"children":748},{},[749],{"type":50,"value":750},"BYO Container path (any framework, any language):",{"type":44,"tag":52,"props":752,"children":753},{},[754,756,762,764,770],{"type":50,"value":755},"For frameworks or languages not scaffolded by the CLI, AgentCore Runtime accepts any container that implements the HTTP contract (",{"type":44,"tag":90,"props":757,"children":759},{"className":758},[],[760],{"type":50,"value":761},"POST \u002Finvocations",{"type":50,"value":763},", ",{"type":44,"tag":90,"props":765,"children":767},{"className":766},[],[768],{"type":50,"value":769},"GET \u002Fping",{"type":50,"value":771},"). The workflow:",{"type":44,"tag":773,"props":774,"children":775},"ol",{},[776,787,798,811],{"type":44,"tag":69,"props":777,"children":778},{},[779,785],{"type":44,"tag":90,"props":780,"children":782},{"className":781},[],[783],{"type":50,"value":784},"agentcore create --name \u003CProjectName> --defaults",{"type":50,"value":786}," to scaffold the project structure",{"type":44,"tag":69,"props":788,"children":789},{},[790,796],{"type":44,"tag":90,"props":791,"children":793},{"className":792},[],[794],{"type":50,"value":795},"agentcore add agent --type byo --build Container --language \u003CLanguage> --code-location \u003Cpath>",{"type":50,"value":797}," to register your code",{"type":44,"tag":69,"props":799,"children":800},{},[801,803,809],{"type":50,"value":802},"Write a ",{"type":44,"tag":90,"props":804,"children":806},{"className":805},[],[807],{"type":50,"value":808},"Dockerfile",{"type":50,"value":810}," that builds and runs your agent",{"type":44,"tag":69,"props":812,"children":813},{},[814,820],{"type":44,"tag":90,"props":815,"children":817},{"className":816},[],[818],{"type":50,"value":819},"agentcore deploy",{"type":50,"value":821}," handles ECR push, CDK infra, and runtime creation",{"type":44,"tag":52,"props":823,"children":824},{},[825],{"type":44,"tag":337,"props":826,"children":827},{},[828],{"type":50,"value":829},"Language-specific notes:",{"type":44,"tag":435,"props":831,"children":832},{},[833,849],{"type":44,"tag":439,"props":834,"children":835},{},[836],{"type":44,"tag":443,"props":837,"children":838},{},[839,844],{"type":44,"tag":447,"props":840,"children":841},{},[842],{"type":50,"value":843},"Language",{"type":44,"tag":447,"props":845,"children":846},{},[847],{"type":50,"value":848},"Recommended path",{"type":44,"tag":463,"props":850,"children":851},{},[852,881,901],{"type":44,"tag":443,"props":853,"children":854},{},[855,860],{"type":44,"tag":470,"props":856,"children":857},{},[858],{"type":50,"value":859},"Java (Spring Boot)",{"type":44,"tag":470,"props":861,"children":862},{},[863,871,873,879],{"type":44,"tag":149,"props":864,"children":868},{"href":865,"rel":866},"https:\u002F\u002Faws.amazon.com\u002Fblogs\u002Fmachine-learning\u002Fspring-ai-sdk-for-amazon-bedrock-agentcore-is-now-generally-available",[867],"nofollow",[869],{"type":50,"value":870},"Spring AI SDK for AgentCore",{"type":50,"value":872}," — handles the Runtime contract, SSE streaming, and health checks. Use ",{"type":44,"tag":90,"props":874,"children":876},{"className":875},[],[877],{"type":50,"value":878},"--language Other --build Container",{"type":50,"value":880},".",{"type":44,"tag":443,"props":882,"children":883},{},[884,889],{"type":44,"tag":470,"props":885,"children":886},{},[887],{"type":50,"value":888},"JavaScript \u002F TypeScript",{"type":44,"tag":470,"props":890,"children":891},{},[892,894,900],{"type":50,"value":893},"Implement the Runtime contract in Express\u002FFastify\u002Fetc. Use ",{"type":44,"tag":90,"props":895,"children":897},{"className":896},[],[898],{"type":50,"value":899},"--language TypeScript --build Container",{"type":50,"value":880},{"type":44,"tag":443,"props":902,"children":903},{},[904,909],{"type":44,"tag":470,"props":905,"children":906},{},[907],{"type":50,"value":908},"Go, Rust, .NET, other",{"type":44,"tag":470,"props":910,"children":911},{},[912,914,919],{"type":50,"value":913},"Implement the Runtime HTTP contract. Use ",{"type":44,"tag":90,"props":915,"children":917},{"className":916},[],[918],{"type":50,"value":878},{"type":50,"value":880},{"type":44,"tag":52,"props":921,"children":922},{},[923],{"type":50,"value":924},"The rest of this skill (deploy, status, logs, invoke) applies once the container builds correctly.",{"type":44,"tag":622,"props":926,"children":928},{"id":927},"framework-vs-model-provider-a-common-confusion",[929],{"type":50,"value":930},"Framework vs. model provider — a common confusion",{"type":44,"tag":52,"props":932,"children":933},{},[934],{"type":50,"value":935},"The framework is how your agent orchestrates (Strands, LangGraph, etc.). The model provider is which LLM it calls (Bedrock, Anthropic, OpenAI, Gemini). These are independent choices:",{"type":44,"tag":65,"props":937,"children":938},{},[939,944,949,954],{"type":44,"tag":69,"props":940,"children":941},{},[942],{"type":50,"value":943},"Strands + Bedrock (default) — AWS-native everything",{"type":44,"tag":69,"props":945,"children":946},{},[947],{"type":50,"value":948},"Strands + Anthropic — Strands orchestration, direct Anthropic API for the model",{"type":44,"tag":69,"props":950,"children":951},{},[952],{"type":50,"value":953},"LangGraph + Bedrock — LangGraph orchestration, Bedrock for the model",{"type":44,"tag":69,"props":955,"children":956},{},[957],{"type":50,"value":958},"OpenAI Agents + OpenAI — OpenAI everything",{"type":44,"tag":52,"props":960,"children":961},{},[962],{"type":50,"value":963},"If the developer says \"I want to use Claude\" they mean the model provider (Bedrock or Anthropic), not the framework. If they say \"I want to use LangGraph\" they mean the framework.",{"type":44,"tag":208,"props":965,"children":967},{"id":966},"step-3-create-the-project",[968],{"type":50,"value":969},"Step 3: Create the project",{"type":44,"tag":52,"props":971,"children":972},{},[973,975,980],{"type":50,"value":974},"Build the ",{"type":44,"tag":90,"props":976,"children":978},{"className":977},[],[979],{"type":50,"value":95},{"type":50,"value":981}," command based on the developer's choices.",{"type":44,"tag":52,"props":983,"children":984},{},[985,990,992,997],{"type":44,"tag":337,"props":986,"children":987},{},[988],{"type":50,"value":989},"Before constructing the command — validate the project name.",{"type":50,"value":991}," The CLI fails late: if the name is invalid, you'll see the error ",{"type":44,"tag":704,"props":993,"children":994},{},[995],{"type":50,"value":996},"after",{"type":50,"value":998}," walking through prompts or building the full command. Save the round-trip and check these rules up front. Reject the name and ask for a new one if any rule fails:",{"type":44,"tag":65,"props":1000,"children":1001},{},[1002,1050,1060],{"type":44,"tag":69,"props":1003,"children":1004},{},[1005,1010,1012,1018,1020,1026,1028,1034,1036,1042,1044,1049],{"type":44,"tag":337,"props":1006,"children":1007},{},[1008],{"type":50,"value":1009},"Length ≤ 23 characters",{"type":50,"value":1011}," (this is shorter than most developers assume — ",{"type":44,"tag":90,"props":1013,"children":1015},{"className":1014},[],[1016],{"type":50,"value":1017},"MyCustomerSupportAgent",{"type":50,"value":1019}," is 22 chars and fits; ",{"type":44,"tag":90,"props":1021,"children":1023},{"className":1022},[],[1024],{"type":50,"value":1025},"CustomerSupportChatbot",{"type":50,"value":1027}," is 22 and fits; ",{"type":44,"tag":90,"props":1029,"children":1031},{"className":1030},[],[1032],{"type":50,"value":1033},"MyCustomerSupportBotApp",{"type":50,"value":1035}," is 23 and just fits; ",{"type":44,"tag":90,"props":1037,"children":1039},{"className":1038},[],[1040],{"type":50,"value":1041},"MyCustomerSupportChatBot",{"type":50,"value":1043}," is 24 and ",{"type":44,"tag":337,"props":1045,"children":1046},{},[1047],{"type":50,"value":1048},"fails",{"type":50,"value":160},{"type":44,"tag":69,"props":1051,"children":1052},{},[1053,1058],{"type":44,"tag":337,"props":1054,"children":1055},{},[1056],{"type":50,"value":1057},"Alphanumeric only",{"type":50,"value":1059}," — no hyphens, underscores, dots, or spaces",{"type":44,"tag":69,"props":1061,"children":1062},{},[1063],{"type":44,"tag":337,"props":1064,"children":1065},{},[1066],{"type":50,"value":1067},"Must start with a letter",{"type":44,"tag":52,"props":1069,"children":1070},{},[1071,1073,1079],{"type":50,"value":1072},"Say the count back out loud when close to the limit: \"That name is 24 characters — the CLI caps project names at 23. Want to shorten it to ",{"type":44,"tag":90,"props":1074,"children":1076},{"className":1075},[],[1077],{"type":50,"value":1078},"\u003Csuggestion>",{"type":50,"value":1080},"?\" Do not run the command with an invalid name on the assumption that the CLI error message will be clear — it isn't always, and the developer's mental model will be wrong for subsequent commands.",{"type":44,"tag":52,"props":1082,"children":1083},{},[1084,1089],{"type":44,"tag":337,"props":1085,"children":1086},{},[1087],{"type":50,"value":1088},"Construct the command, then present it for confirmation before the developer runs it.",{"type":50,"value":1090}," Show the full command with all flags and explain what each choice means. Wait for the developer to confirm or adjust before proceeding.",{"type":44,"tag":52,"props":1092,"children":1093},{},[1094],{"type":50,"value":1095},"Example presentation:",{"type":44,"tag":254,"props":1097,"children":1098},{},[1099,1104,1173,1178],{"type":44,"tag":52,"props":1100,"children":1101},{},[1102],{"type":50,"value":1103},"Here's the command I'd recommend based on what you've told me:",{"type":44,"tag":215,"props":1105,"children":1107},{"className":217,"code":1106,"language":219,"meta":220,"style":220},"agentcore create --name MyAgent --framework Strands --model-provider Bedrock --build CodeZip --memory none\n",[1108],{"type":44,"tag":90,"props":1109,"children":1110},{"__ignoreMap":220},[1111],{"type":44,"tag":226,"props":1112,"children":1113},{"class":228,"line":229},[1114,1118,1123,1128,1133,1138,1143,1148,1153,1158,1163,1168],{"type":44,"tag":226,"props":1115,"children":1116},{"style":233},[1117],{"type":50,"value":236},{"type":44,"tag":226,"props":1119,"children":1120},{"style":239},[1121],{"type":50,"value":1122}," create",{"type":44,"tag":226,"props":1124,"children":1125},{"style":239},[1126],{"type":50,"value":1127}," --name",{"type":44,"tag":226,"props":1129,"children":1130},{"style":239},[1131],{"type":50,"value":1132}," MyAgent",{"type":44,"tag":226,"props":1134,"children":1135},{"style":239},[1136],{"type":50,"value":1137}," --framework",{"type":44,"tag":226,"props":1139,"children":1140},{"style":239},[1141],{"type":50,"value":1142}," Strands",{"type":44,"tag":226,"props":1144,"children":1145},{"style":239},[1146],{"type":50,"value":1147}," --model-provider",{"type":44,"tag":226,"props":1149,"children":1150},{"style":239},[1151],{"type":50,"value":1152}," Bedrock",{"type":44,"tag":226,"props":1154,"children":1155},{"style":239},[1156],{"type":50,"value":1157}," --build",{"type":44,"tag":226,"props":1159,"children":1160},{"style":239},[1161],{"type":50,"value":1162}," CodeZip",{"type":44,"tag":226,"props":1164,"children":1165},{"style":239},[1166],{"type":50,"value":1167}," --memory",{"type":44,"tag":226,"props":1169,"children":1170},{"style":239},[1171],{"type":50,"value":1172}," none\n",{"type":44,"tag":52,"props":1174,"children":1175},{},[1176],{"type":50,"value":1177},"This creates a Strands agent using Bedrock models, deployed as a code zip (no Docker needed). Memory can be added later.",{"type":44,"tag":52,"props":1179,"children":1180},{},[1181],{"type":50,"value":1182},"Want to run this, or change anything?",{"type":44,"tag":52,"props":1184,"children":1185},{},[1186],{"type":50,"value":1187},"Do NOT execute the command automatically — present it and wait.",{"type":44,"tag":52,"props":1189,"children":1190},{},[1191],{"type":44,"tag":337,"props":1192,"children":1193},{},[1194],{"type":50,"value":1195},"Minimal (defaults — Strands, Bedrock, CodeZip, no memory):",{"type":44,"tag":215,"props":1197,"children":1199},{"className":217,"code":1198,"language":219,"meta":220,"style":220},"agentcore create --name \u003CProjectName> --defaults\n",[1200],{"type":44,"tag":90,"props":1201,"children":1202},{"__ignoreMap":220},[1203],{"type":44,"tag":226,"props":1204,"children":1205},{"class":228,"line":229},[1206,1210,1214,1218,1224,1229,1235,1240],{"type":44,"tag":226,"props":1207,"children":1208},{"style":233},[1209],{"type":50,"value":236},{"type":44,"tag":226,"props":1211,"children":1212},{"style":239},[1213],{"type":50,"value":1122},{"type":44,"tag":226,"props":1215,"children":1216},{"style":239},[1217],{"type":50,"value":1127},{"type":44,"tag":226,"props":1219,"children":1221},{"style":1220},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[1222],{"type":50,"value":1223}," \u003C",{"type":44,"tag":226,"props":1225,"children":1226},{"style":239},[1227],{"type":50,"value":1228},"ProjectNam",{"type":44,"tag":226,"props":1230,"children":1232},{"style":1231},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[1233],{"type":50,"value":1234},"e",{"type":44,"tag":226,"props":1236,"children":1237},{"style":1220},[1238],{"type":50,"value":1239},">",{"type":44,"tag":226,"props":1241,"children":1242},{"style":239},[1243],{"type":50,"value":1244}," --defaults\n",{"type":44,"tag":52,"props":1246,"children":1247},{},[1248],{"type":44,"tag":337,"props":1249,"children":1250},{},[1251],{"type":50,"value":1252},"With specific options:",{"type":44,"tag":215,"props":1254,"children":1256},{"className":217,"code":1255,"language":219,"meta":220,"style":220},"agentcore create \\\n  --name \u003CProjectName> \\\n  --framework \u003CFramework> \\\n  --model-provider Bedrock \\\n  --build CodeZip \\\n  --memory none\n",[1257],{"type":44,"tag":90,"props":1258,"children":1259},{"__ignoreMap":220},[1260,1276,1305,1336,1353,1370],{"type":44,"tag":226,"props":1261,"children":1262},{"class":228,"line":229},[1263,1267,1271],{"type":44,"tag":226,"props":1264,"children":1265},{"style":233},[1266],{"type":50,"value":236},{"type":44,"tag":226,"props":1268,"children":1269},{"style":239},[1270],{"type":50,"value":1122},{"type":44,"tag":226,"props":1272,"children":1273},{"style":1231},[1274],{"type":50,"value":1275}," \\\n",{"type":44,"tag":226,"props":1277,"children":1279},{"class":228,"line":1278},2,[1280,1285,1289,1293,1297,1301],{"type":44,"tag":226,"props":1281,"children":1282},{"style":239},[1283],{"type":50,"value":1284},"  --name",{"type":44,"tag":226,"props":1286,"children":1287},{"style":1220},[1288],{"type":50,"value":1223},{"type":44,"tag":226,"props":1290,"children":1291},{"style":239},[1292],{"type":50,"value":1228},{"type":44,"tag":226,"props":1294,"children":1295},{"style":1231},[1296],{"type":50,"value":1234},{"type":44,"tag":226,"props":1298,"children":1299},{"style":1220},[1300],{"type":50,"value":1239},{"type":44,"tag":226,"props":1302,"children":1303},{"style":1231},[1304],{"type":50,"value":1275},{"type":44,"tag":226,"props":1306,"children":1308},{"class":228,"line":1307},3,[1309,1314,1318,1323,1328,1332],{"type":44,"tag":226,"props":1310,"children":1311},{"style":239},[1312],{"type":50,"value":1313},"  --framework",{"type":44,"tag":226,"props":1315,"children":1316},{"style":1220},[1317],{"type":50,"value":1223},{"type":44,"tag":226,"props":1319,"children":1320},{"style":239},[1321],{"type":50,"value":1322},"Framewor",{"type":44,"tag":226,"props":1324,"children":1325},{"style":1231},[1326],{"type":50,"value":1327},"k",{"type":44,"tag":226,"props":1329,"children":1330},{"style":1220},[1331],{"type":50,"value":1239},{"type":44,"tag":226,"props":1333,"children":1334},{"style":1231},[1335],{"type":50,"value":1275},{"type":44,"tag":226,"props":1337,"children":1339},{"class":228,"line":1338},4,[1340,1345,1349],{"type":44,"tag":226,"props":1341,"children":1342},{"style":239},[1343],{"type":50,"value":1344},"  --model-provider",{"type":44,"tag":226,"props":1346,"children":1347},{"style":239},[1348],{"type":50,"value":1152},{"type":44,"tag":226,"props":1350,"children":1351},{"style":1231},[1352],{"type":50,"value":1275},{"type":44,"tag":226,"props":1354,"children":1356},{"class":228,"line":1355},5,[1357,1362,1366],{"type":44,"tag":226,"props":1358,"children":1359},{"style":239},[1360],{"type":50,"value":1361},"  --build",{"type":44,"tag":226,"props":1363,"children":1364},{"style":239},[1365],{"type":50,"value":1162},{"type":44,"tag":226,"props":1367,"children":1368},{"style":1231},[1369],{"type":50,"value":1275},{"type":44,"tag":226,"props":1371,"children":1373},{"class":228,"line":1372},6,[1374,1379],{"type":44,"tag":226,"props":1375,"children":1376},{"style":239},[1377],{"type":50,"value":1378},"  --memory",{"type":44,"tag":226,"props":1380,"children":1381},{"style":239},[1382],{"type":50,"value":1172},{"type":44,"tag":52,"props":1384,"children":1385},{},[1386],{"type":44,"tag":337,"props":1387,"children":1388},{},[1389],{"type":50,"value":1390},"Flag reference:",{"type":44,"tag":435,"props":1392,"children":1393},{},[1394,1415],{"type":44,"tag":439,"props":1395,"children":1396},{},[1397],{"type":44,"tag":443,"props":1398,"children":1399},{},[1400,1405,1410],{"type":44,"tag":447,"props":1401,"children":1402},{},[1403],{"type":50,"value":1404},"Flag",{"type":44,"tag":447,"props":1406,"children":1407},{},[1408],{"type":50,"value":1409},"Values",{"type":44,"tag":447,"props":1411,"children":1412},{},[1413],{"type":50,"value":1414},"Default",{"type":44,"tag":463,"props":1416,"children":1417},{},[1418,1440,1481,1524,1560,1606,1645,1681],{"type":44,"tag":443,"props":1419,"children":1420},{},[1421,1430,1435],{"type":44,"tag":470,"props":1422,"children":1423},{},[1424],{"type":44,"tag":90,"props":1425,"children":1427},{"className":1426},[],[1428],{"type":50,"value":1429},"--name",{"type":44,"tag":470,"props":1431,"children":1432},{},[1433],{"type":50,"value":1434},"alphanumeric, max 23 chars",{"type":44,"tag":470,"props":1436,"children":1437},{},[1438],{"type":50,"value":1439},"prompted",{"type":44,"tag":443,"props":1441,"children":1442},{},[1443,1451,1477],{"type":44,"tag":470,"props":1444,"children":1445},{},[1446],{"type":44,"tag":90,"props":1447,"children":1449},{"className":1448},[],[1450],{"type":50,"value":582},{"type":44,"tag":470,"props":1452,"children":1453},{},[1454,1459,1460,1465,1466,1471,1472],{"type":44,"tag":90,"props":1455,"children":1457},{"className":1456},[],[1458],{"type":50,"value":474},{"type":50,"value":763},{"type":44,"tag":90,"props":1461,"children":1463},{"className":1462},[],[1464],{"type":50,"value":504},{"type":50,"value":763},{"type":44,"tag":90,"props":1467,"children":1469},{"className":1468},[],[1470],{"type":50,"value":526},{"type":50,"value":763},{"type":44,"tag":90,"props":1473,"children":1475},{"className":1474},[],[1476],{"type":50,"value":548},{"type":44,"tag":470,"props":1478,"children":1479},{},[1480],{"type":50,"value":1439},{"type":44,"tag":443,"props":1482,"children":1483},{},[1484,1493,1516],{"type":44,"tag":470,"props":1485,"children":1486},{},[1487],{"type":44,"tag":90,"props":1488,"children":1490},{"className":1489},[],[1491],{"type":50,"value":1492},"--protocol",{"type":44,"tag":470,"props":1494,"children":1495},{},[1496,1502,1503,1509,1510],{"type":44,"tag":90,"props":1497,"children":1499},{"className":1498},[],[1500],{"type":50,"value":1501},"HTTP",{"type":50,"value":763},{"type":44,"tag":90,"props":1504,"children":1506},{"className":1505},[],[1507],{"type":50,"value":1508},"MCP",{"type":50,"value":763},{"type":44,"tag":90,"props":1511,"children":1513},{"className":1512},[],[1514],{"type":50,"value":1515},"A2A",{"type":44,"tag":470,"props":1517,"children":1518},{},[1519],{"type":44,"tag":90,"props":1520,"children":1522},{"className":1521},[],[1523],{"type":50,"value":1501},{"type":44,"tag":443,"props":1525,"children":1526},{},[1527,1536,1552],{"type":44,"tag":470,"props":1528,"children":1529},{},[1530],{"type":44,"tag":90,"props":1531,"children":1533},{"className":1532},[],[1534],{"type":50,"value":1535},"--build",{"type":44,"tag":470,"props":1537,"children":1538},{},[1539,1545,1546],{"type":44,"tag":90,"props":1540,"children":1542},{"className":1541},[],[1543],{"type":50,"value":1544},"CodeZip",{"type":50,"value":763},{"type":44,"tag":90,"props":1547,"children":1549},{"className":1548},[],[1550],{"type":50,"value":1551},"Container",{"type":44,"tag":470,"props":1553,"children":1554},{},[1555],{"type":44,"tag":90,"props":1556,"children":1558},{"className":1557},[],[1559],{"type":50,"value":1544},{"type":44,"tag":443,"props":1561,"children":1562},{},[1563,1572,1602],{"type":44,"tag":470,"props":1564,"children":1565},{},[1566],{"type":44,"tag":90,"props":1567,"children":1569},{"className":1568},[],[1570],{"type":50,"value":1571},"--model-provider",{"type":44,"tag":470,"props":1573,"children":1574},{},[1575,1581,1582,1588,1589,1595,1596],{"type":44,"tag":90,"props":1576,"children":1578},{"className":1577},[],[1579],{"type":50,"value":1580},"Bedrock",{"type":50,"value":763},{"type":44,"tag":90,"props":1583,"children":1585},{"className":1584},[],[1586],{"type":50,"value":1587},"Anthropic",{"type":50,"value":763},{"type":44,"tag":90,"props":1590,"children":1592},{"className":1591},[],[1593],{"type":50,"value":1594},"OpenAI",{"type":50,"value":763},{"type":44,"tag":90,"props":1597,"children":1599},{"className":1598},[],[1600],{"type":50,"value":1601},"Gemini",{"type":44,"tag":470,"props":1603,"children":1604},{},[1605],{"type":50,"value":1439},{"type":44,"tag":443,"props":1607,"children":1608},{},[1609,1618,1641],{"type":44,"tag":470,"props":1610,"children":1611},{},[1612],{"type":44,"tag":90,"props":1613,"children":1615},{"className":1614},[],[1616],{"type":50,"value":1617},"--memory",{"type":44,"tag":470,"props":1619,"children":1620},{},[1621,1627,1628,1634,1635],{"type":44,"tag":90,"props":1622,"children":1624},{"className":1623},[],[1625],{"type":50,"value":1626},"none",{"type":50,"value":763},{"type":44,"tag":90,"props":1629,"children":1631},{"className":1630},[],[1632],{"type":50,"value":1633},"shortTerm",{"type":50,"value":763},{"type":44,"tag":90,"props":1636,"children":1638},{"className":1637},[],[1639],{"type":50,"value":1640},"longAndShortTerm",{"type":44,"tag":470,"props":1642,"children":1643},{},[1644],{"type":50,"value":1439},{"type":44,"tag":443,"props":1646,"children":1647},{},[1648,1657,1673],{"type":44,"tag":470,"props":1649,"children":1650},{},[1651],{"type":44,"tag":90,"props":1652,"children":1654},{"className":1653},[],[1655],{"type":50,"value":1656},"--network-mode",{"type":44,"tag":470,"props":1658,"children":1659},{},[1660,1666,1667],{"type":44,"tag":90,"props":1661,"children":1663},{"className":1662},[],[1664],{"type":50,"value":1665},"PUBLIC",{"type":50,"value":763},{"type":44,"tag":90,"props":1668,"children":1670},{"className":1669},[],[1671],{"type":50,"value":1672},"VPC",{"type":44,"tag":470,"props":1674,"children":1675},{},[1676],{"type":44,"tag":90,"props":1677,"children":1679},{"className":1678},[],[1680],{"type":50,"value":1665},{"type":44,"tag":443,"props":1682,"children":1683},{},[1684,1693,1698],{"type":44,"tag":470,"props":1685,"children":1686},{},[1687],{"type":44,"tag":90,"props":1688,"children":1690},{"className":1689},[],[1691],{"type":50,"value":1692},"--dry-run",{"type":44,"tag":470,"props":1694,"children":1695},{},[1696],{"type":50,"value":1697},"—",{"type":44,"tag":470,"props":1699,"children":1700},{},[1701],{"type":50,"value":1702},"preview without creating",{"type":44,"tag":52,"props":1704,"children":1705},{},[1706],{"type":44,"tag":337,"props":1707,"children":1708},{},[1709],{"type":50,"value":1710},"Guidance on choices:",{"type":44,"tag":65,"props":1712,"children":1713},{},[1714,1731,1747,1763],{"type":44,"tag":69,"props":1715,"children":1716},{},[1717,1722,1724,1729],{"type":44,"tag":337,"props":1718,"children":1719},{},[1720],{"type":50,"value":1721},"Protocol:",{"type":50,"value":1723}," Use ",{"type":44,"tag":90,"props":1725,"children":1727},{"className":1726},[],[1728],{"type":50,"value":1501},{"type":50,"value":1730}," unless the developer specifically needs MCP tool serving or A2A agent-to-agent communication",{"type":44,"tag":69,"props":1732,"children":1733},{},[1734,1739,1740,1745],{"type":44,"tag":337,"props":1735,"children":1736},{},[1737],{"type":50,"value":1738},"Build:",{"type":50,"value":1723},{"type":44,"tag":90,"props":1741,"children":1743},{"className":1742},[],[1744],{"type":50,"value":1544},{"type":50,"value":1746}," unless the developer needs custom system dependencies (CodeZip is faster to deploy and doesn't require Docker locally)",{"type":44,"tag":69,"props":1748,"children":1749},{},[1750,1755,1756,1761],{"type":44,"tag":337,"props":1751,"children":1752},{},[1753],{"type":50,"value":1754},"Model provider:",{"type":50,"value":1723},{"type":44,"tag":90,"props":1757,"children":1759},{"className":1758},[],[1760],{"type":50,"value":1580},{"type":50,"value":1762}," unless the developer has a specific reason for another provider (Bedrock doesn't require managing API keys)",{"type":44,"tag":69,"props":1764,"children":1765},{},[1766,1771,1773,1778,1780,1785,1786,1796],{"type":44,"tag":337,"props":1767,"children":1768},{},[1769],{"type":50,"value":1770},"Memory:",{"type":50,"value":1772}," Start with ",{"type":44,"tag":90,"props":1774,"children":1776},{"className":1775},[],[1777],{"type":50,"value":1626},{"type":50,"value":1779}," — memory can be added later via ",{"type":44,"tag":90,"props":1781,"children":1783},{"className":1782},[],[1784],{"type":50,"value":127},{"type":50,"value":147},{"type":44,"tag":149,"props":1787,"children":1789},{"href":1788},"..\u002Fagents-build\u002Freferences\u002Fmemory.md",[1790],{"type":44,"tag":90,"props":1791,"children":1793},{"className":1792},[],[1794],{"type":50,"value":1795},"references\u002Fmemory.md",{"type":50,"value":1797},") when the developer needs it",{"type":44,"tag":208,"props":1799,"children":1801},{"id":1800},"step-4-explain-what-was-created",[1802],{"type":50,"value":1803},"Step 4: Explain what was created",{"type":44,"tag":52,"props":1805,"children":1806},{},[1807,1809,1814],{"type":50,"value":1808},"After the project exists, read ",{"type":44,"tag":90,"props":1810,"children":1812},{"className":1811},[],[1813],{"type":50,"value":390},{"type":50,"value":1815}," and the generated code to explain the project structure.",{"type":44,"tag":52,"props":1817,"children":1818},{},[1819,1821,1827],{"type":50,"value":1820},"The layout below reflects CLI v0.9.x. If the CLI version is different, run ",{"type":44,"tag":90,"props":1822,"children":1824},{"className":1823},[],[1825],{"type":50,"value":1826},"tree \u003CProjectName>\u002F -L 3",{"type":50,"value":1828}," to see the actual generated structure and explain from there.",{"type":44,"tag":215,"props":1830,"children":1834},{"className":1831,"code":1833,"language":50},[1832],"language-text","\u003CProjectName>\u002F\n├── agentcore\u002F\n│   ├── agentcore.json      ← Project config (agents, resources)\n│   ├── aws-targets.json    ← AWS account + region\n│   ├── .env.local          ← Local environment variables (gitignored)\n│   └── cdk\u002F                ← CDK infrastructure (auto-managed, don't edit)\n└── app\u002F\n    └── \u003CAgentName>\u002F\n        ├── main.py          ← Your agent code — this is where you build\n        ├── mcp_client\u002F      ← Pre-wired example MCP client (see note below)\n        └── pyproject.toml   ← Python dependencies\n",[1835],{"type":44,"tag":90,"props":1836,"children":1837},{"__ignoreMap":220},[1838],{"type":50,"value":1833},{"type":44,"tag":52,"props":1840,"children":1841},{},[1842],{"type":44,"tag":337,"props":1843,"children":1844},{},[1845],{"type":50,"value":1846},"Key files to highlight:",{"type":44,"tag":65,"props":1848,"children":1849},{},[1850,1861,1879],{"type":44,"tag":69,"props":1851,"children":1852},{},[1853,1859],{"type":44,"tag":90,"props":1854,"children":1856},{"className":1855},[],[1857],{"type":50,"value":1858},"app\u002F\u003CAgentName>\u002Fmain.py",{"type":50,"value":1860}," — the agent's entry point. This is where the developer adds tools, system prompts, and logic.",{"type":44,"tag":69,"props":1862,"children":1863},{},[1864,1869,1871,1877],{"type":44,"tag":90,"props":1865,"children":1867},{"className":1866},[],[1868],{"type":50,"value":390},{"type":50,"value":1870}," — the project config. Resources are added here via ",{"type":44,"tag":90,"props":1872,"children":1874},{"className":1873},[],[1875],{"type":50,"value":1876},"agentcore add",{"type":50,"value":1878}," commands.",{"type":44,"tag":69,"props":1880,"children":1881},{},[1882,1888],{"type":44,"tag":90,"props":1883,"children":1885},{"className":1884},[],[1886],{"type":50,"value":1887},"agentcore\u002F.env.local",{"type":50,"value":1889}," — local environment variables. After deploy, resource IDs are written here for local dev.",{"type":44,"tag":52,"props":1891,"children":1892},{},[1893,1898,1900,1906,1908,1914,1916,1922,1924,1930,1932,1938],{"type":44,"tag":337,"props":1894,"children":1895},{},[1896],{"type":50,"value":1897},"Heads-up on the scaffolded MCP client.",{"type":50,"value":1899}," ",{"type":44,"tag":90,"props":1901,"children":1903},{"className":1902},[],[1904],{"type":50,"value":1905},"main.py",{"type":50,"value":1907}," imports ",{"type":44,"tag":90,"props":1909,"children":1911},{"className":1910},[],[1912],{"type":50,"value":1913},"get_streamable_http_mcp_client()",{"type":50,"value":1915}," from ",{"type":44,"tag":90,"props":1917,"children":1919},{"className":1918},[],[1920],{"type":50,"value":1921},"mcp_client\u002Fclient.py",{"type":50,"value":1923}," and appends it to ",{"type":44,"tag":90,"props":1925,"children":1927},{"className":1926},[],[1928],{"type":50,"value":1929},"tools",{"type":50,"value":1931},". In a fresh project, this client points at a public example MCP endpoint — so ",{"type":44,"tag":90,"props":1933,"children":1935},{"className":1934},[],[1936],{"type":50,"value":1937},"agentcore dev",{"type":50,"value":1939}," works immediately. Two things to flag:",{"type":44,"tag":773,"props":1941,"children":1942},{},[1943,2004],{"type":44,"tag":69,"props":1944,"children":1945},{},[1946,1951,1953,1959,1961,1966,1968,1973,1975,1980,1982,1988,1990,1995,1997,2003],{"type":44,"tag":337,"props":1947,"children":1948},{},[1949],{"type":50,"value":1950},"It will become a silent no-op if you repoint it at a gateway that isn't deployed yet.",{"type":50,"value":1952}," The common path is to swap the example endpoint for ",{"type":44,"tag":90,"props":1954,"children":1956},{"className":1955},[],[1957],{"type":50,"value":1958},"os.getenv(\"AGENTCORE_GATEWAY_\u003CNAME>_URL\")",{"type":50,"value":1960},". That env var is only populated after ",{"type":44,"tag":90,"props":1962,"children":1964},{"className":1963},[],[1965],{"type":50,"value":819},{"type":50,"value":1967},". If the developer repoints and runs ",{"type":44,"tag":90,"props":1969,"children":1971},{"className":1970},[],[1972],{"type":50,"value":1937},{"type":50,"value":1974}," before deploying, ",{"type":44,"tag":90,"props":1976,"children":1978},{"className":1977},[],[1979],{"type":50,"value":1913},{"type":50,"value":1981}," returns a client with a ",{"type":44,"tag":90,"props":1983,"children":1985},{"className":1984},[],[1986],{"type":50,"value":1987},"None",{"type":50,"value":1989}," URL and the agent starts with zero MCP tools — no error, no warning. See the \"Local dev gap\" section in ",{"type":44,"tag":90,"props":1991,"children":1993},{"className":1992},[],[1994],{"type":50,"value":135},{"type":50,"value":1996}," for the guard pattern: ",{"type":44,"tag":90,"props":1998,"children":2000},{"className":1999},[],[2001],{"type":50,"value":2002},"if not GATEWAY_URL: tools = []",{"type":50,"value":880},{"type":44,"tag":69,"props":2005,"children":2006},{},[2007,2012,2014,2020,2022,2027],{"type":44,"tag":337,"props":2008,"children":2009},{},[2010],{"type":50,"value":2011},"If the developer doesn't need MCP tools at all",{"type":50,"value":2013},", remove the ",{"type":44,"tag":90,"props":2015,"children":2017},{"className":2016},[],[2018],{"type":50,"value":2019},"mcp_clients",{"type":50,"value":2021}," list and the loop that appends it to ",{"type":44,"tag":90,"props":2023,"children":2025},{"className":2024},[],[2026],{"type":50,"value":1929},{"type":50,"value":2028},". The scaffold includes it as a convenience, not a requirement.",{"type":44,"tag":52,"props":2030,"children":2031},{},[2032,2034,2039],{"type":50,"value":2033},"The reference client code in ",{"type":44,"tag":90,"props":2035,"children":2037},{"className":2036},[],[2038],{"type":50,"value":135},{"type":50,"value":2040}," (Path A) shows the correct pattern for gateway-backed MCP clients once deploy has run.",{"type":44,"tag":208,"props":2042,"children":2044},{"id":2043},"step-5-local-development",[2045],{"type":50,"value":2046},"Step 5: Local development",{"type":44,"tag":215,"props":2048,"children":2050},{"className":217,"code":2049,"language":219,"meta":220,"style":220},"agentcore dev\n",[2051],{"type":44,"tag":90,"props":2052,"children":2053},{"__ignoreMap":220},[2054],{"type":44,"tag":226,"props":2055,"children":2056},{"class":228,"line":229},[2057,2061],{"type":44,"tag":226,"props":2058,"children":2059},{"style":233},[2060],{"type":50,"value":236},{"type":44,"tag":226,"props":2062,"children":2063},{"style":239},[2064],{"type":50,"value":2065}," dev\n",{"type":44,"tag":52,"props":2067,"children":2068},{},[2069],{"type":50,"value":2070},"This starts a local dev server. The developer can interact with their agent immediately.",{"type":44,"tag":52,"props":2072,"children":2073},{},[2074,2079,2081,2087],{"type":44,"tag":337,"props":2075,"children":2076},{},[2077],{"type":50,"value":2078},"Port the dev server binds to",{"type":50,"value":2080}," (important if you're scripting ",{"type":44,"tag":90,"props":2082,"children":2084},{"className":2083},[],[2085],{"type":50,"value":2086},"curl",{"type":50,"value":2088}," calls or testing from another process):",{"type":44,"tag":435,"props":2090,"children":2091},{},[2092,2108],{"type":44,"tag":439,"props":2093,"children":2094},{},[2095],{"type":44,"tag":443,"props":2096,"children":2097},{},[2098,2103],{"type":44,"tag":447,"props":2099,"children":2100},{},[2101],{"type":50,"value":2102},"Protocol",{"type":44,"tag":447,"props":2104,"children":2105},{},[2106],{"type":50,"value":2107},"Default port",{"type":44,"tag":463,"props":2109,"children":2110},{},[2111,2127,2143],{"type":44,"tag":443,"props":2112,"children":2113},{},[2114,2118],{"type":44,"tag":470,"props":2115,"children":2116},{},[2117],{"type":50,"value":1501},{"type":44,"tag":470,"props":2119,"children":2120},{},[2121],{"type":44,"tag":90,"props":2122,"children":2124},{"className":2123},[],[2125],{"type":50,"value":2126},"8080",{"type":44,"tag":443,"props":2128,"children":2129},{},[2130,2134],{"type":44,"tag":470,"props":2131,"children":2132},{},[2133],{"type":50,"value":1508},{"type":44,"tag":470,"props":2135,"children":2136},{},[2137],{"type":44,"tag":90,"props":2138,"children":2140},{"className":2139},[],[2141],{"type":50,"value":2142},"8000",{"type":44,"tag":443,"props":2144,"children":2145},{},[2146,2150],{"type":44,"tag":470,"props":2147,"children":2148},{},[2149],{"type":50,"value":1515},{"type":44,"tag":470,"props":2151,"children":2152},{},[2153],{"type":44,"tag":90,"props":2154,"children":2156},{"className":2155},[],[2157],{"type":50,"value":2158},"9000",{"type":44,"tag":52,"props":2160,"children":2161},{},[2162,2164,2169,2171,2177,2179,2185],{"type":50,"value":2163},"The CLI prints the bound port and URL on startup — always read the actual value from the CLI output rather than hardcoding. ",{"type":44,"tag":337,"props":2165,"children":2166},{},[2167],{"type":50,"value":2168},"If the default port is already in use",{"type":50,"value":2170},", the CLI auto-increments (e.g., 8080 → 8081 → 8082), so a second dev session or a lingering process from a previous run can shift your port without warning. Use ",{"type":44,"tag":90,"props":2172,"children":2174},{"className":2173},[],[2175],{"type":50,"value":2176},"agentcore dev --port \u003CN>",{"type":50,"value":2178}," to pin it, or grep ",{"type":44,"tag":90,"props":2180,"children":2182},{"className":2181},[],[2183],{"type":50,"value":2184},"ps",{"type":50,"value":2186}," \u002F check the CLI banner if invocations start failing with connection-refused or exit-code-7 errors.",{"type":44,"tag":52,"props":2188,"children":2189},{},[2190],{"type":44,"tag":337,"props":2191,"children":2192},{},[2193],{"type":50,"value":2194},"Important limitations to mention:",{"type":44,"tag":65,"props":2196,"children":2197},{},[2198,2210,2215],{"type":44,"tag":69,"props":2199,"children":2200},{},[2201,2203,2208],{"type":50,"value":2202},"Memory is not available in ",{"type":44,"tag":90,"props":2204,"children":2206},{"className":2205},[],[2207],{"type":50,"value":1937},{"type":50,"value":2209}," — it requires a deploy",{"type":44,"tag":69,"props":2211,"children":2212},{},[2213],{"type":50,"value":2214},"Gateway URLs are not available locally — they require a deploy",{"type":44,"tag":69,"props":2216,"children":2217},{},[2218],{"type":50,"value":2219},"The local server uses the model provider configured in the project",{"type":44,"tag":208,"props":2221,"children":2223},{"id":2222},"step-6-first-deploy",[2224],{"type":50,"value":2225},"Step 6: First deploy",{"type":44,"tag":52,"props":2227,"children":2228},{},[2229],{"type":50,"value":2230},"When the developer is ready to deploy:",{"type":44,"tag":215,"props":2232,"children":2234},{"className":217,"code":2233,"language":219,"meta":220,"style":220},"agentcore deploy\n",[2235],{"type":44,"tag":90,"props":2236,"children":2237},{"__ignoreMap":220},[2238],{"type":44,"tag":226,"props":2239,"children":2240},{"class":228,"line":229},[2241,2245],{"type":44,"tag":226,"props":2242,"children":2243},{"style":233},[2244],{"type":50,"value":236},{"type":44,"tag":226,"props":2246,"children":2247},{"style":239},[2248],{"type":50,"value":2249}," deploy\n",{"type":44,"tag":52,"props":2251,"children":2252},{},[2253],{"type":50,"value":2254},"This will:",{"type":44,"tag":773,"props":2256,"children":2257},{},[2258,2263,2268],{"type":44,"tag":69,"props":2259,"children":2260},{},[2261],{"type":50,"value":2262},"Show a preview of AWS resources to be created",{"type":44,"tag":69,"props":2264,"children":2265},{},[2266],{"type":50,"value":2267},"Ask for confirmation",{"type":44,"tag":69,"props":2269,"children":2270},{},[2271],{"type":50,"value":2272},"Build and deploy via CDK",{"type":44,"tag":52,"props":2274,"children":2275},{},[2276,2281],{"type":44,"tag":337,"props":2277,"children":2278},{},[2279],{"type":50,"value":2280},"First deploy takes 3-5 minutes.",{"type":50,"value":2282}," Subsequent deploys are faster.",{"type":44,"tag":52,"props":2284,"children":2285},{},[2286],{"type":50,"value":2287},"After deploy, show them how to invoke:",{"type":44,"tag":215,"props":2289,"children":2291},{"className":217,"code":2290,"language":219,"meta":220,"style":220},"agentcore invoke \"Hello, what can you do?\"\n",[2292],{"type":44,"tag":90,"props":2293,"children":2294},{"__ignoreMap":220},[2295],{"type":44,"tag":226,"props":2296,"children":2297},{"class":228,"line":229},[2298,2302,2307,2312,2317],{"type":44,"tag":226,"props":2299,"children":2300},{"style":233},[2301],{"type":50,"value":236},{"type":44,"tag":226,"props":2303,"children":2304},{"style":239},[2305],{"type":50,"value":2306}," invoke",{"type":44,"tag":226,"props":2308,"children":2309},{"style":1220},[2310],{"type":50,"value":2311}," \"",{"type":44,"tag":226,"props":2313,"children":2314},{"style":239},[2315],{"type":50,"value":2316},"Hello, what can you do?",{"type":44,"tag":226,"props":2318,"children":2319},{"style":1220},[2320],{"type":50,"value":2321},"\"\n",{"type":44,"tag":52,"props":2323,"children":2324},{},[2325],{"type":50,"value":2326},"And how to check status:",{"type":44,"tag":215,"props":2328,"children":2330},{"className":217,"code":2329,"language":219,"meta":220,"style":220},"agentcore status\n",[2331],{"type":44,"tag":90,"props":2332,"children":2333},{"__ignoreMap":220},[2334],{"type":44,"tag":226,"props":2335,"children":2336},{"class":228,"line":229},[2337,2341],{"type":44,"tag":226,"props":2338,"children":2339},{"style":233},[2340],{"type":50,"value":236},{"type":44,"tag":226,"props":2342,"children":2343},{"style":239},[2344],{"type":50,"value":2345}," status\n",{"type":44,"tag":208,"props":2347,"children":2349},{"id":2348},"step-7-whats-next",[2350],{"type":50,"value":2351},"Step 7: What's next",{"type":44,"tag":52,"props":2353,"children":2354},{},[2355],{"type":50,"value":2356},"Based on what the developer said they want to build, suggest the logical next skill:",{"type":44,"tag":435,"props":2358,"children":2359},{},[2360,2381],{"type":44,"tag":439,"props":2361,"children":2362},{},[2363],{"type":44,"tag":443,"props":2364,"children":2365},{},[2366,2371,2376],{"type":44,"tag":447,"props":2367,"children":2368},{},[2369],{"type":50,"value":2370},"Developer intent",{"type":44,"tag":447,"props":2372,"children":2373},{},[2374],{"type":50,"value":2375},"Next skill",{"type":44,"tag":447,"props":2377,"children":2378},{},[2379],{"type":50,"value":2380},"Command hint",{"type":44,"tag":463,"props":2382,"children":2383},{},[2384,2409,2434,2459,2484,2510,2532,2557],{"type":44,"tag":443,"props":2385,"children":2386},{},[2387,2392,2400],{"type":44,"tag":470,"props":2388,"children":2389},{},[2390],{"type":50,"value":2391},"\"How do I call it from my app?\"",{"type":44,"tag":470,"props":2393,"children":2394},{},[2395],{"type":44,"tag":90,"props":2396,"children":2398},{"className":2397},[],[2399],{"type":50,"value":127},{"type":44,"tag":470,"props":2401,"children":2402},{},[2403],{"type":44,"tag":90,"props":2404,"children":2406},{"className":2405},[],[2407],{"type":50,"value":2408},"agentcore fetch access",{"type":44,"tag":443,"props":2410,"children":2411},{},[2412,2417,2425],{"type":44,"tag":470,"props":2413,"children":2414},{},[2415],{"type":50,"value":2416},"\"I want it to remember things\"",{"type":44,"tag":470,"props":2418,"children":2419},{},[2420],{"type":44,"tag":90,"props":2421,"children":2423},{"className":2422},[],[2424],{"type":50,"value":127},{"type":44,"tag":470,"props":2426,"children":2427},{},[2428],{"type":44,"tag":90,"props":2429,"children":2431},{"className":2430},[],[2432],{"type":50,"value":2433},"agentcore add memory",{"type":44,"tag":443,"props":2435,"children":2436},{},[2437,2442,2450],{"type":44,"tag":470,"props":2438,"children":2439},{},[2440],{"type":50,"value":2441},"\"I want it to call external APIs\"",{"type":44,"tag":470,"props":2443,"children":2444},{},[2445],{"type":44,"tag":90,"props":2446,"children":2448},{"className":2447},[],[2449],{"type":50,"value":135},{"type":44,"tag":470,"props":2451,"children":2452},{},[2453],{"type":44,"tag":90,"props":2454,"children":2456},{"className":2455},[],[2457],{"type":50,"value":2458},"agentcore add gateway",{"type":44,"tag":443,"props":2460,"children":2461},{},[2462,2467,2475],{"type":44,"tag":470,"props":2463,"children":2464},{},[2465],{"type":50,"value":2466},"\"I want to restrict what it can do\"",{"type":44,"tag":470,"props":2468,"children":2469},{},[2470],{"type":44,"tag":90,"props":2471,"children":2473},{"className":2472},[],[2474],{"type":50,"value":135},{"type":44,"tag":470,"props":2476,"children":2477},{},[2478],{"type":44,"tag":90,"props":2479,"children":2481},{"className":2480},[],[2482],{"type":50,"value":2483},"agentcore add policy-engine",{"type":44,"tag":443,"props":2485,"children":2486},{},[2487,2492,2501],{"type":44,"tag":470,"props":2488,"children":2489},{},[2490],{"type":50,"value":2491},"\"I want to measure quality\"",{"type":44,"tag":470,"props":2493,"children":2494},{},[2495],{"type":44,"tag":90,"props":2496,"children":2498},{"className":2497},[],[2499],{"type":50,"value":2500},"agents-optimize",{"type":44,"tag":470,"props":2502,"children":2503},{},[2504],{"type":44,"tag":90,"props":2505,"children":2507},{"className":2506},[],[2508],{"type":50,"value":2509},"agentcore add evaluator",{"type":44,"tag":443,"props":2511,"children":2512},{},[2513,2518,2527],{"type":44,"tag":470,"props":2514,"children":2515},{},[2516],{"type":50,"value":2517},"\"I want to go to production\"",{"type":44,"tag":470,"props":2519,"children":2520},{},[2521],{"type":44,"tag":90,"props":2522,"children":2524},{"className":2523},[],[2525],{"type":50,"value":2526},"agents-harden",{"type":44,"tag":470,"props":2528,"children":2529},{},[2530],{"type":50,"value":2531},"production readiness checklist",{"type":44,"tag":443,"props":2533,"children":2534},{},[2535,2540,2548],{"type":44,"tag":470,"props":2536,"children":2537},{},[2538],{"type":50,"value":2539},"\"I want multiple agents working together\"",{"type":44,"tag":470,"props":2541,"children":2542},{},[2543],{"type":44,"tag":90,"props":2544,"children":2546},{"className":2545},[],[2547],{"type":50,"value":127},{"type":44,"tag":470,"props":2549,"children":2550},{},[2551],{"type":44,"tag":90,"props":2552,"children":2554},{"className":2553},[],[2555],{"type":50,"value":2556},"agentcore create --protocol A2A",{"type":44,"tag":443,"props":2558,"children":2559},{},[2560,2565,2573],{"type":44,"tag":470,"props":2561,"children":2562},{},[2563],{"type":50,"value":2564},"\"I need it in a VPC\"",{"type":44,"tag":470,"props":2566,"children":2567},{},[2568],{"type":44,"tag":90,"props":2569,"children":2571},{"className":2570},[],[2572],{"type":50,"value":127},{"type":44,"tag":470,"props":2574,"children":2575},{},[2576],{"type":44,"tag":90,"props":2577,"children":2579},{"className":2578},[],[2580],{"type":50,"value":2581},"agentcore create --network-mode VPC",{"type":44,"tag":52,"props":2583,"children":2584},{},[2585],{"type":50,"value":2586},"Don't overwhelm — suggest one or two next steps based on what the developer actually asked for.",{"type":44,"tag":208,"props":2588,"children":2590},{"id":2589},"example-walkthroughs",[2591],{"type":50,"value":2592},"Example walkthroughs",{"type":44,"tag":52,"props":2594,"children":2595},{},[2596],{"type":50,"value":2597},"For task-framed prompts (e.g., \"build a customer support agent\"), load the matching example reference:",{"type":44,"tag":435,"props":2599,"children":2600},{},[2601,2617],{"type":44,"tag":439,"props":2602,"children":2603},{},[2604],{"type":44,"tag":443,"props":2605,"children":2606},{},[2607,2612],{"type":44,"tag":447,"props":2608,"children":2609},{},[2610],{"type":50,"value":2611},"Developer task",{"type":44,"tag":447,"props":2613,"children":2614},{},[2615],{"type":50,"value":2616},"Reference",{"type":44,"tag":463,"props":2618,"children":2619},{},[2620],{"type":44,"tag":443,"props":2621,"children":2622},{},[2623,2628],{"type":44,"tag":470,"props":2624,"children":2625},{},[2626],{"type":50,"value":2627},"Customer support, chatbot, answer policy questions",{"type":44,"tag":470,"props":2629,"children":2630},{},[2631],{"type":44,"tag":149,"props":2632,"children":2634},{"href":2633},"references\u002Fexample-support-agent.md",[2635],{"type":44,"tag":90,"props":2636,"children":2638},{"className":2637},[],[2639],{"type":50,"value":2633},{"type":44,"tag":52,"props":2641,"children":2642},{},[2643],{"type":50,"value":2644},"More examples can be added to this skill's references directory as common patterns emerge.",{"type":44,"tag":58,"props":2646,"children":2648},{"id":2647},"output",[2649],{"type":50,"value":2650},"Output",{"type":44,"tag":65,"props":2652,"children":2653},{},[2654,2659,2671,2676],{"type":44,"tag":69,"props":2655,"children":2656},{},[2657],{"type":50,"value":2658},"A clear path from \"I want to build an agent\" to a running deployed agent",{"type":44,"tag":69,"props":2660,"children":2661},{},[2662,2664,2669],{"type":50,"value":2663},"The ",{"type":44,"tag":90,"props":2665,"children":2667},{"className":2666},[],[2668],{"type":50,"value":95},{"type":50,"value":2670}," command tailored to their choices",{"type":44,"tag":69,"props":2672,"children":2673},{},[2674],{"type":50,"value":2675},"An explanation of the generated project structure",{"type":44,"tag":69,"props":2677,"children":2678},{},[2679],{"type":50,"value":2680},"Concrete next steps based on their intent",{"type":44,"tag":58,"props":2682,"children":2684},{"id":2683},"quality-criteria",[2685],{"type":50,"value":2686},"Quality criteria",{"type":44,"tag":65,"props":2688,"children":2689},{},[2690,2701,2706,2711],{"type":44,"tag":69,"props":2691,"children":2692},{},[2693,2694,2699],{"type":50,"value":2663},{"type":44,"tag":90,"props":2695,"children":2697},{"className":2696},[],[2698],{"type":50,"value":95},{"type":50,"value":2700}," command uses only valid flags from CLI v0.9.1",{"type":44,"tag":69,"props":2702,"children":2703},{},[2704],{"type":50,"value":2705},"Framework recommendation is based on the developer's context, not a generic default",{"type":44,"tag":69,"props":2707,"children":2708},{},[2709],{"type":50,"value":2710},"The developer understands what each generated file does",{"type":44,"tag":69,"props":2712,"children":2713},{},[2714],{"type":50,"value":2715},"Next steps are specific to what the developer wants to build, not a generic list of all features",{"type":44,"tag":2717,"props":2718,"children":2719},"style",{},[2720],{"type":50,"value":2721},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":2723,"total":2814},[2724,2738,2752,2766,2779,2785,2799],{"slug":127,"name":127,"fn":2725,"description":2726,"org":2727,"tags":2728,"stars":21,"repoUrl":22,"updatedAt":2737},"add capabilities to existing agent projects","Use when adding capabilities to an existing agent project — memory, app integration, VPC, multi-agent, migration, model changes, browser, code interpreter, or resource removal. Triggers on: \"add memory\", \"remember across sessions\", \"call agent from app\", \"invoke agent from code\", \"auth to call agent\", \"streaming responses\", \"VPC\", \"VPC connectivity\", \"VPC error\", \"can't reach from VPC\", \"multi-agent\", \"A2A\", \"A2A auth\", \"orchestrator not delegating\", \"specialist not called\", \"migrate Bedrock Agent\", \"after import\", \"migration issue\", \"framework for migration\", \"change model\", \"browser tool\", \"code interpreter\", \"delete agent\", \"tear down\", \"agentcore remove\", \"cross-account memory\", \"resource-based policy on memory\", \"pay for x402 content\", \"402 Payment Required\", \"microtransactions\", \"paid API or tool\". Not for connecting to external APIs via Gateway — use agents-connect. Not for scaffolding a new project — use agents-get-started. Not for CLI\u002Fdev server errors — use agents-debug. Strands vs LangGraph in a migration context routes here.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2729,2730,2733,2734],{"name":17,"slug":18,"type":15},{"name":2731,"slug":2732,"type":15},"Automation","automation",{"name":20,"slug":8,"type":15},{"name":2735,"slug":2736,"type":15},"Engineering","engineering","2026-07-12T08:42:53.812877",{"slug":135,"name":135,"fn":2739,"description":2740,"org":2741,"tags":2742,"stars":21,"repoUrl":22,"updatedAt":2751},"connect agents to external services","Use when connecting your agent to external APIs, tools, or services via Gateway, or restricting tool access with Cedar policies. Handles gateway setup, target types, outbound auth (OAuth, API key, IAM), credentials, and Cedar policy authoring. Triggers on: \"connect to API\", \"add gateway\", \"connect to MCP server\", \"Lambda tools\", \"OpenAPI\", \"gateway target\", \"Cedar policy\", \"restrict tools\", \"policy engine\", \"gateway auth error\", \"store API key\", \"outbound credential\", \"env var API key\", \"API key None after deploy\", \"credential not available after deploy\", \"should this be a gateway target\", \"give my agent tools\", \"add tools to agent\". Not for inbound auth (who can call your agent) — use agents-harden. Not for debugging agent behavior — use agents-debug. Not for VPC networking errors (agent can't reach APIs due to VPC) — use agents-build. Not for creating or hosting a new MCP server project — use agents-get-started.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2743,2744,2747,2750],{"name":17,"slug":18,"type":15},{"name":2745,"slug":2746,"type":15},"API Development","api-development",{"name":2748,"slug":2749,"type":15},"Authentication","authentication",{"name":20,"slug":8,"type":15},"2026-07-16T06:00:38.866147",{"slug":116,"name":116,"fn":2753,"description":2754,"org":2755,"tags":2756,"stars":21,"repoUrl":22,"updatedAt":2765},"debug agent and environment issues","Use when your agent or environment is broken — wrong answers, errors, timeouts, tool failures, or CLI issues. Reads traces and logs to diagnose root causes. Also checks prerequisites when the CLI itself isn't working. Triggers on: \"agent not working\", \"wrong answer\", \"agent error\", \"tool call failing\", \"debug agent\", \"check logs\", \"read traces\", \"broken\", \"500 error\", \"424 error\", \"model access denied\", \"command not found\", \"stuck in DELETING\", \"maxVms exceeded\", \"cold start diagnosis\", \"cold start slow\", \"agentcore create error\", \"create failed\", \"exit code 7\", \"connection refused local dev\". Not for deploy failures — use agents-deploy. Not for performance tuning without errors — use agents-optimize. Not for VPC configuration — use agents-build. Not for observability setup or missing logs — use agents-optimize.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2757,2758,2759,2762],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2760,"slug":2761,"type":15},"Debugging","debugging",{"name":2763,"slug":2764,"type":15},"Observability","observability","2026-07-16T06:00:44.679093",{"slug":2767,"name":2767,"fn":2768,"description":2769,"org":2770,"tags":2771,"stars":21,"repoUrl":22,"updatedAt":2778},"agents-deploy","deploy AI agents to AWS","Use when deploying your agent to AWS, or when a deploy has failed. Handles pre-flight validation, CDK\u002FIAM\u002Fquota error diagnosis, version management, rollback, and canary deployments. Triggers on: \"deploy my agent\", \"agentcore deploy\", \"deploy failed\", \"CDK error\", \"rollback\", \"canary deploy\", \"pin version\", \"redeploy\", \"deploy stuck\". Not for production hardening — use agents-harden. Not for adding capabilities before deploy — use agents-build or agents-connect. Not for VPC configuration errors — use agents-build.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2772,2773,2774,2777],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2775,"slug":2776,"type":15},"CI\u002FCD","ci-cd",{"name":13,"slug":14,"type":15},"2026-07-12T08:42:55.059577",{"slug":4,"name":4,"fn":5,"description":6,"org":2780,"tags":2781,"stars":21,"repoUrl":22,"updatedAt":23},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2782,2783,2784],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":13,"slug":14,"type":15},{"slug":2526,"name":2526,"fn":2786,"description":2787,"org":2788,"tags":2789,"stars":21,"repoUrl":22,"updatedAt":2798},"harden agents for production","Use when preparing your agent for production — IAM scoping, inbound auth (JWT, SigV4), secrets management, cold start optimization, session lifecycle, rate limiting, input validation, and quota guidance. Triggers on: \"production checklist\", \"harden agent\", \"production ready\", \"secure agent\", \"inbound auth\", \"going live\", \"cold start optimization\", \"session lifecycle\", \"StopRuntimeSession\", \"quota\", \"throttling\", \"maxVms\", \"rate limit\", \"security audit of outbound API calls\", \"gateway target audit for production\", \"restrict who can call\", \"lock down endpoint\", \"only our app can call\". Not for Cedar tool-restriction policies — use agents-connect. Not for quality measurement — use agents-optimize. Not for outbound credential storage or API key wiring — use agents-connect. Not for A2A agent-to-agent auth — use agents-build. Cold start observation and diagnosis (not optimization) routes to agents-debug.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2790,2791,2792,2795],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2793,"slug":2794,"type":15},"Best Practices","best-practices",{"name":2796,"slug":2797,"type":15},"Security","security","2026-07-16T06:00:42.174705",{"slug":2500,"name":2500,"fn":2800,"description":2801,"org":2802,"tags":2803,"stars":21,"repoUrl":22,"updatedAt":2813},"optimize agent quality and performance","Use when measuring or improving agent quality and performance — set up evaluators, online monitoring, CI\u002FCD quality gates, observability, or cost optimization. Triggers on: \"evaluate my agent\", \"add evaluator\", \"measure quality\", \"quality gate\", \"run evals\", \"agent too slow\", \"why is it slow\", \"reduce latency\", \"set up observability\", \"CloudWatch dashboard\", \"how much does my agent cost\", \"cost optimization\", \"logs not showing up\", \"logs missing\", \"spans not found\", \"eval failing\", \"eval error\", \"dev traces\", \"local traces\", \"agentcore dev traces\", \"traces to CloudWatch\". Not for debugging errors or crashes — use agents-debug. Slow but correct routes here; broken routes to debug.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2804,2805,2806,2809,2810],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2807,"slug":2808,"type":15},"Evals","evals",{"name":2763,"slug":2764,"type":15},{"name":2811,"slug":2812,"type":15},"Performance","performance","2026-07-12T08:42:56.488105",114,{"items":2816,"total":2933},[2817,2824,2831,2838,2845,2851,2858,2866,2883,2896,2908,2923],{"slug":127,"name":127,"fn":2725,"description":2726,"org":2818,"tags":2819,"stars":21,"repoUrl":22,"updatedAt":2737},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2820,2821,2822,2823],{"name":17,"slug":18,"type":15},{"name":2731,"slug":2732,"type":15},{"name":20,"slug":8,"type":15},{"name":2735,"slug":2736,"type":15},{"slug":135,"name":135,"fn":2739,"description":2740,"org":2825,"tags":2826,"stars":21,"repoUrl":22,"updatedAt":2751},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2827,2828,2829,2830],{"name":17,"slug":18,"type":15},{"name":2745,"slug":2746,"type":15},{"name":2748,"slug":2749,"type":15},{"name":20,"slug":8,"type":15},{"slug":116,"name":116,"fn":2753,"description":2754,"org":2832,"tags":2833,"stars":21,"repoUrl":22,"updatedAt":2765},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2834,2835,2836,2837],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2760,"slug":2761,"type":15},{"name":2763,"slug":2764,"type":15},{"slug":2767,"name":2767,"fn":2768,"description":2769,"org":2839,"tags":2840,"stars":21,"repoUrl":22,"updatedAt":2778},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2841,2842,2843,2844],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2775,"slug":2776,"type":15},{"name":13,"slug":14,"type":15},{"slug":4,"name":4,"fn":5,"description":6,"org":2846,"tags":2847,"stars":21,"repoUrl":22,"updatedAt":23},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2848,2849,2850],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":13,"slug":14,"type":15},{"slug":2526,"name":2526,"fn":2786,"description":2787,"org":2852,"tags":2853,"stars":21,"repoUrl":22,"updatedAt":2798},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2854,2855,2856,2857],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2793,"slug":2794,"type":15},{"name":2796,"slug":2797,"type":15},{"slug":2500,"name":2500,"fn":2800,"description":2801,"org":2859,"tags":2860,"stars":21,"repoUrl":22,"updatedAt":2813},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2861,2862,2863,2864,2865],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2807,"slug":2808,"type":15},{"name":2763,"slug":2764,"type":15},{"name":2811,"slug":2812,"type":15},{"slug":2867,"name":2867,"fn":2868,"description":2869,"org":2870,"tags":2871,"stars":21,"repoUrl":22,"updatedAt":2882},"amazon-aurora-mysql","manage Amazon Aurora MySQL clusters","Amazon Aurora MySQL — creates, modifies, and advises on Aurora MySQL clusters specifically (MySQL-compatible engine, Aurora serverless, parallel query). Trigger for Aurora MySQL cluster operations, ACU sizing, I\u002FO-Optimized storage, commitment pricing, or MySQL upgrade planning. Aurora MySQL uses full (VPC-based) configuration — express configuration is PostgreSQL-only. For Aurora PostgreSQL, use amazon-aurora-postgresql instead. Contains safety guardrails and response templates that override defaults.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2872,2873,2876,2879],{"name":20,"slug":8,"type":15},{"name":2874,"slug":2875,"type":15},"Database","database",{"name":2877,"slug":2878,"type":15},"MySQL","mysql",{"name":2880,"slug":2881,"type":15},"Serverless","serverless","2026-07-12T08:43:13.27939",{"slug":2884,"name":2884,"fn":2885,"description":2886,"org":2887,"tags":2888,"stars":21,"repoUrl":22,"updatedAt":2895},"amazon-aurora-postgresql","configure Amazon Aurora PostgreSQL clusters","Amazon Aurora PostgreSQL — creates, modifies, and advises on Aurora PostgreSQL clusters specifically (PostgreSQL-compatible engine, Aurora serverless, express configuration, pgvector, Babelfish). Trigger for Aurora PostgreSQL cluster operations, express-configuration quick-start, ACU sizing, I\u002FO-Optimized storage, commitment pricing, or PostgreSQL upgrade planning. For Aurora MySQL, use amazon-aurora-mysql instead. Contains safety guardrails, express-first routing, and response templates that override defaults.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2889,2890,2891,2894],{"name":20,"slug":8,"type":15},{"name":2874,"slug":2875,"type":15},{"name":2892,"slug":2893,"type":15},"PostgreSQL","postgresql",{"name":2880,"slug":2881,"type":15},"2026-07-16T06:00:34.789624",{"slug":2897,"name":2897,"fn":2898,"description":2899,"org":2900,"tags":2901,"stars":21,"repoUrl":22,"updatedAt":2907},"amazon-bedrock","build generative AI apps with Amazon Bedrock","Builds generative AI applications on Amazon Bedrock. Covers model invocation (Converse API, InvokeModel), RAG with Knowledge Bases, Bedrock Agents, Guardrails, and AgentCore. Use when invoking models, setting up Knowledge Bases, creating agents, applying guardrails, deploying to AgentCore, migrating\u002Fporting\u002Fconverting a Bedrock Agent (including inline agents) to an AgentCore Harness, troubleshooting Bedrock errors (ThrottlingException, AccessDeniedException), or choosing models (Claude, Llama, Nova, Titan). ALSO USE for prompt caching setup and debugging, quota health checks and throttling diagnosis, cost attribution and tracking, migrating between Claude model generations (4.5 to 4.6 to 4.7), chunking strategies, API selection (Converse vs InvokeModel), guardrail capabilities, and model selection. Also covers AgentCore Payments setup (x402, microtransactions, Payment Manager, Connector, Instrument, Coinbase CDP, Stripe Privy, 402 Payment Required, pay for content, paid endpoint, agent payments). NOT for custom model training, Rekognition, or Comprehend.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2902,2903,2904],{"name":17,"slug":18,"type":15},{"name":20,"slug":8,"type":15},{"name":2905,"slug":2906,"type":15},"LLM","llm","2026-07-25T05:30:35.20899",{"slug":2909,"name":2909,"fn":2910,"description":2911,"org":2912,"tags":2913,"stars":21,"repoUrl":22,"updatedAt":2922},"amazon-documentdb","manage Amazon DocumentDB clusters","Manages Amazon DocumentDB end-to-end — serverless-on-8.0 cluster setup, TLS\u002FVPC\u002Fdriver config, flexible-schema and vector-search data modeling, MongoDB compatibility assessment, DMS-based migration, slow-query diagnosis, major version upgrades (4.0→5.0→8.0), Well-Architected reviews (41-check wa_review.py), cost estimation, and security hardening. Retrieve for every DocumentDB question and when the user asks to set up or migrate MongoDB to AWS — DocumentDB is AWS's MongoDB-compatible managed database. Triggers: JSON document store, document database, MongoDB on AWS, Nested fields, Lambda cannot connect, TLS handshake, VPC port 27017, IAM auth, Secrets Manager, encryption at rest, $graphLookup, flexible schema, COLLSCAN, compound index, DMS migration, CDC cutover, $vectorSearch, RAG, Global Clusters, DR replication, cost sizing, audit, health check, production-readiness.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2914,2915,2916,2919],{"name":20,"slug":8,"type":15},{"name":2874,"slug":2875,"type":15},{"name":2917,"slug":2918,"type":15},"MongoDB","mongodb",{"name":2920,"slug":2921,"type":15},"NoSQL","nosql","2026-07-12T08:43:00.455878",{"slug":2924,"name":2924,"fn":2925,"description":2926,"org":2927,"tags":2928,"stars":21,"repoUrl":22,"updatedAt":2932},"amazon-dynamodb","design and debug DynamoDB data layers","Designs, reviews, and debugs DynamoDB data layers from design axioms — enumerates access patterns, chooses partition\u002Fsort keys and GSIs, decides single-table vs. multi-table, configures Streams, Global Tables, TTL, and zero-ETL integrations to OpenSearch\u002FRedshift\u002FSageMaker Lakehouse, and produces a defensible data-layer design with a monthly cost estimate and optional live validation. Applies whenever a user is designing, reviewing, or refactoring anything backed by DynamoDB — schemas, access patterns, GSIs, single- vs. multi-table choices, Streams consumers, transactional outboxes, Global Tables, zero-ETL pipelines — even when they don't say \"axioms\" or \"design review.\" Also applies when debugging hot partitions, throttling, unbounded Scans, LWW conflicts, or surprise bills on DynamoDB workloads.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2929,2930,2931],{"name":20,"slug":8,"type":15},{"name":2874,"slug":2875,"type":15},{"name":2920,"slug":2921,"type":15},"2026-07-16T06:00:37.690386",115]