[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-cline-desktop-commander-overview":3,"mdc--rokd6m-key":42,"related-org-cline-desktop-commander-overview":1233,"related-repo-cline-desktop-commander-overview":1433},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":31,"repoUrl":32,"updatedAt":33,"license":34,"forks":35,"topics":36,"repo":37,"sourceUrl":40,"mdContent":41},"desktop-commander-overview","manage local desktop processes and files","Use for Desktop Commander MCP capabilities — persistent shells and REPLs, long-running processes, filesystem beyond the workspace, structured files (.xlsx, .docx, .pdf, images) and large local data files such as CSVs, ripgrep search at scale, SSH, or cross-turn state.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"cline","Cline","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fcline.png",[12,16,19,22,25,28],{"name":13,"slug":14,"type":15},"PDF","pdf","tag",{"name":17,"slug":18,"type":15},"Excel","excel",{"name":20,"slug":21,"type":15},"Automation","automation",{"name":23,"slug":24,"type":15},"MCP","mcp",{"name":26,"slug":27,"type":15},"Desktop","desktop",{"name":29,"slug":30,"type":15},"DOCX","docx",10,"https:\u002F\u002Fgithub.com\u002Fcline\u002Fskills","2026-07-12T08:13:58.245511",null,4,[],{"repoUrl":32,"stars":31,"forks":35,"topics":38,"description":39},[],"A collection of skills used at Cline","https:\u002F\u002Fgithub.com\u002Fcline\u002Fskills\u002Ftree\u002FHEAD\u002Fskills\u002Fdesktop-commander-overview","---\nname: desktop-commander-overview\ndescription: Use for Desktop Commander MCP capabilities — persistent shells and REPLs, long-running processes, filesystem beyond the workspace, structured files (.xlsx, .docx, .pdf, images) and large local data files such as CSVs, ripgrep search at scale, SSH, or cross-turn state.\nversion: 0.1.0\naudience: agent\n---\n\n# Desktop Commander MCP\n\nDesktop Commander gives the agent reach across the user's actual computer — files, folders, terminals, processes, structured documents, and remote machines reachable over SSH. The tools' detailed schemas (parameters, return shapes, format-specific behavior) live in the MCP itself; this skill explains what they enable and how they compose into common workflows.\n\n## What this MCP gives the agent\n\n**Persistent shell sessions.** Desktop Commander keeps a started process or session alive across tool calls. Inside a single long-lived shell, REPL, or SSH session, state carries forward — environment variables, working directory, activated virtualenvs, open connections, REPL variables — so the agent can `cd`, activate a venv, then send commands or code into that same session many turns later without re-setup. (Note: separate `start_process` calls open separate sessions and do **not** share shell state with each other; persistence is inside one session, not across them.)\n\n**Long-running processes.** Start a dev server, watcher, build, training run, or test suite in the background and keep working. The MCP returns a process handle the agent can tail, interact with, or terminate across many turns. Long-running commands don't need to block the workflow waiting for a foreground command to exit.\n\n**Filesystem reach beyond the IDE workspace.** Read, write, move, list, and inspect files anywhere the user has granted scope — Downloads, Documents, project folders outside the IDE, or any other granted folders. Useful for organize-and-clean tasks, batch document work, and any \"look at the file my coworker just sent me\" request that doesn't fit inside the IDE sandbox.\n\n**Surgical edits to existing files.** The `edit_block` tool does exact-string find-and-replace with built-in safety: ambiguous matches fail loudly instead of silently overwriting the wrong thing, and an `expected_replacements` count prevents partial-match disasters. Lower data-loss risk than rewriting whole files based on the slice you happened to read — though a wrong `old_string` or wrong `expected_replacements` can still corrupt content, so review the changed content before considering the edit done.\n\n**Binary and structured files handled directly by the MCP.** Excel, DOCX, and PDF are first-class — read and modified through format-specific mechanisms rather than text-only approximations: Excel via cell-range JSON, DOCX via raw-XML edits, PDF via page-level operations on a new output file. The result is the real file in its original format, not a regenerated approximation. Images and PDFs return as viewable content for the agent.\n\n**Search at scale.** Streaming, ripgrep-backed search across whole projects or folder trees. The agent picks between filename search and in-file content search, pages through results progressively without flooding context, and runs multiple concurrent searches when the query is ambiguous.\n\n**Remote machines via SSH.** A long-lived SSH session inside a persistent shell turns the agent into a real ops tool: connect once, then tail logs, run diagnostics, deploy, or debug across many turns without reconnecting each step.\n\n**Process management.** List, inspect, tail, and kill accessible processes (subject to OS permissions). Useful for cleaning up stale dev servers from previous sessions and for diagnosing CPU \u002F memory issues.\n\n## Example workflows\n\nEach example names the actual tool sequence. Calls below are written in pseudocode shorthand (`tool_name(\"arg\", flag=value)`); the real tools take object-shaped arguments. Tool descriptions and full parameter sets live in the MCP itself.\n\n### \"Debug this production issue\"\n\nBefore running production-impacting SSH commands, explain the intended action and get user confirmation when the risk is non-trivial.\n\n`start_process(\"ssh user@prod.example.com\", timeout_ms=...)` opens a long-lived SSH session and returns a PID. `interact_with_process(pid, \"tail -f \u002Fvar\u002Flog\u002Fapp.log\\n\")` starts streaming logs. Subsequent turns: `read_process_output(pid, offset=-50)` to see the last 50 lines as they arrive, `interact_with_process(pid, \"...\")` to run diagnostic commands in the same session. `force_terminate(pid)` to close the session when done — for sessions opened by `start_process`, `force_terminate` is the correct cleanup tool; `kill_process` is for arbitrary OS PIDs found via `list_processes`.\n\n### \"Deploy this to staging\"\n\nBefore deploys, restarts, migrations, or other environment-changing commands, summarize the action and confirm with the user unless they already explicitly asked for that exact operation.\n\n`start_process` for the deploy command (could be a script, an SSH-piped command, or `kubectl`\u002F`gh` etc.). `read_process_output` to track output and surface errors. If the deploy needs an interactive confirmation, `interact_with_process(pid, \"yes\\n\")`. The session stays alive while the agent watches for completion or rollback.\n\n### \"Run the dev server and iterate on the API\"\n\n`start_process(\"npm run dev\", timeout_ms=...)` keeps the server up. The agent then loops: `edit_block` on the route file, `read_process_output(pid, offset=-30)` to see the server's reload, `start_process(\"curl -s http:\u002F\u002Flocalhost:3000\u002Fapi\u002F...\")` for a one-shot test, repeat. The dev server never has to restart between code changes.\n\n### \"Refactor across this monorepo\"\n\n`start_search(pattern=\"oldFunctionName\", path=repo_root, searchType=\"content\")` scopes every call site. `get_more_search_results(sessionId)` pages through. `read_multiple_files(paths=[...])` confirms ambiguous hits in context. `edit_block(file_path, old_string, new_string)` per site, with `expected_replacements` set when the same substring legitimately appears multiple times in one file. Verify by re-running `start_search` on the old name and paging the results with `get_more_search_results(sessionId)` until the run completes — only then can you confirm zero remaining hits.\n\n### \"Update the Q3 numbers in this spreadsheet and tweak the summary in the report\"\n\n`read_file(path=\"\u002F...\u002Fq3.xlsx\", sheet=\"Revenue\", range=\"A1:F50\")` returns the existing numbers as a JSON 2D array. `edit_block(file_path=\"\u002F...\u002Fq3.xlsx\", range=\"Revenue!C12:C24\", content=[[12345], ...])` updates the cells in place. For the report, DOCX editing is a two-read flow: first `read_file(path=\"\u002F...\u002Freport.docx\")` (offset 0) returns the document's outline (headings + paragraph text) so you can locate the summary section. Then `read_file(path=\"\u002F...\u002Freport.docx\", offset=N, length=...)` with **`N > 0`** returns the raw underlying XML around that section — a non-zero offset is what flips the read into XML mode. Copy an XML fragment from that output as `old_string` and call `edit_block(file_path, old_string, new_string)` with the rewritten XML. The user gets back real `.xlsx` and `.docx` files, not regenerated approximations.\n\n### \"Generate the Q3 report as a PDF\"\n\nCompose markdown content (header, table, charts via embedded HTML), then call `write_pdf` to render it to a new PDF file. The MCP's `write_pdf` tool description specifies the exact parameters and filename rules — follow that.\n\n### \"Insert a cover page into this PDF\"\n\n`write_pdf` also supports modifying existing PDFs via an operations array (insert \u002F delete pages). Use it for existing-PDF edits that produce a new PDF — adding a cover page, removing a section, merging in content from another file. See the `write_pdf` tool description for the operation shapes and parameter rules.\n\n### \"Analyze this 200MB CSV\"\n\n`start_process(\"python3 -i\", timeout_ms=...)` opens a Python REPL and returns a PID. `interact_with_process(pid, \"import pandas as pd; df = pd.read_csv('\u002Fabs\u002Fpath.csv')\")` loads it once. Every subsequent question — `df.describe()`, `df.groupby('col').size()`, plot a chart — runs in the same already-loaded REPL. Libraries don't re-import, the dataframe doesn't re-load. The MCP itself recommends this workflow for any local data-file analysis.\n\n### \"Run a quick Node script\"\n\n`start_process(\"node:local\", timeout_ms=...)` opens a stateless Node execution mode on the MCP server itself — ES imports supported. `start_process` opens the runner; each piece of JS is sent via `interact_with_process(pid, \"\u003Cyour JS here>\")` and runs independently (no shared state between calls). Good for one-shot transformations where keeping a long-lived REPL alive isn't worth it. Don't try to put code into the `start_process` command argument — only the runner type (`node:local`) goes there.\n\n### \"Explain this codebase\"\n\n`list_directory(path=repo_root, depth=3)` for shape. `start_search(pattern=\"export \", path=repo_root, searchType=\"content\")` to find the public surface. `read_multiple_files(paths=[entrypoints])` for the actual code. The agent can keep narrowing without re-asking the user where to look.\n\n### \"Organize my Downloads folder\"\n\nResolve the path to absolute first (e.g., `\u002FUsers\u002F\u003Cuser>\u002FDownloads`, not `~\u002FDownloads`). Then `list_directory(path=\"\u002FUsers\u002F\u003Cuser>\u002FDownloads\", depth=1)` to see what's there. `start_search(pattern=\"*.pdf\", path=\"\u002FUsers\u002F\u003Cuser>\u002FDownloads\", searchType=\"files\")` and similar for other types. `create_directory` for new folders. `move_file` per item. Preview the move plan before executing destructive ops.\n\n### \"Onboard me — what was happening last session?\"\n\n`get_recent_tool_calls(maxResults=200)` returns recent activity with arguments and outputs. `list_sessions` shows still-running terminal sessions. `list_searches` shows in-flight searches. `list_processes` shows what's still alive. Together they reconstruct the work without asking the user to recap.\n\n### \"Why isn't the REPL responding?\"\n\n`list_sessions` — if `Blocked: true`, the REPL is waiting for input rather than hung. `read_process_output(pid, offset=-100)` to see what it last printed (often a prompt). `interact_with_process(pid, \"\u003Cthe input it's waiting for>\\n\")` unblocks it.\n\n## Core tool inventory\n\nGrouped index of the tools an agent reaches for most often. Not exhaustive — the MCP exposes additional config \u002F diagnostics \u002F feedback tools beyond this list. Detailed parameters and return shapes for every tool are in the MCP's own tool descriptions.\n\n- **Process \u002F shell:** `start_process`, `interact_with_process`, `read_process_output`, `list_processes`, `list_sessions`, `kill_process`, `force_terminate`\n- **Files (read\u002Fwrite):** `read_file`, `read_multiple_files`, `write_file`, `edit_block`, `write_pdf`\n- **Filesystem:** `list_directory`, `get_file_info`, `move_file`, `create_directory`\n- **Search:** `start_search`, `get_more_search_results`, `list_searches`, `stop_search`\n- **Diagnostics \u002F config:** `get_recent_tool_calls`, `get_config`\n\n## Conventions\n\n**Prefer absolute paths.** Relative paths may fail depending on the working directory, and tilde paths (`~\u002F...`) may not expand in all contexts. Absolute paths are the most reliable; pass them whenever you can.\n\n**Allowed-directory scope.** File operations only work inside the user's configured `allowedDirectories`. Expect `[DENIED]` markers in `list_directory` output and rejections from `read_file` \u002F `write_file` when the path is out of scope. Surface the rejected path to the user — don't retry.\n\n**When running on macOS:** default shell is zsh. Use `python3` not `python`. Some GNU tools have prefixed names (`gsed` for GNU sed). `brew` is the typical package manager. `open` opens files \u002F apps from the terminal, `mdfind` is the fastest path to exact-filename search via Spotlight. Detect the host platform via `get_config` (or by inspecting `process.platform` \u002F `uname` from a shell) before assuming any of the above — Windows and Linux hosts behave differently.\n\n**Pagination.** Long outputs (file reads, process output, search results) all support `offset` and `length`. Negative offsets read from the end (tail mode). Use these instead of dumping huge results into context.\n",{"data":43,"body":46},{"name":4,"description":6,"version":44,"audience":45},"0.1.0","agent",{"type":47,"children":48},"root",[49,58,64,71,106,116,126,167,177,187,197,207,213,226,233,238,312,318,323,365,371,405,411,468,474,550,556,576,582,599,605,639,645,686,692,719,725,778,784,818,824,858,864,869,1057,1063,1081,1128,1208],{"type":50,"tag":51,"props":52,"children":54},"element","h1",{"id":53},"desktop-commander-mcp",[55],{"type":56,"value":57},"text","Desktop Commander MCP",{"type":50,"tag":59,"props":60,"children":61},"p",{},[62],{"type":56,"value":63},"Desktop Commander gives the agent reach across the user's actual computer — files, folders, terminals, processes, structured documents, and remote machines reachable over SSH. The tools' detailed schemas (parameters, return shapes, format-specific behavior) live in the MCP itself; this skill explains what they enable and how they compose into common workflows.",{"type":50,"tag":65,"props":66,"children":68},"h2",{"id":67},"what-this-mcp-gives-the-agent",[69],{"type":56,"value":70},"What this MCP gives the agent",{"type":50,"tag":59,"props":72,"children":73},{},[74,80,82,89,91,97,99,104],{"type":50,"tag":75,"props":76,"children":77},"strong",{},[78],{"type":56,"value":79},"Persistent shell sessions.",{"type":56,"value":81}," Desktop Commander keeps a started process or session alive across tool calls. Inside a single long-lived shell, REPL, or SSH session, state carries forward — environment variables, working directory, activated virtualenvs, open connections, REPL variables — so the agent can ",{"type":50,"tag":83,"props":84,"children":86},"code",{"className":85},[],[87],{"type":56,"value":88},"cd",{"type":56,"value":90},", activate a venv, then send commands or code into that same session many turns later without re-setup. (Note: separate ",{"type":50,"tag":83,"props":92,"children":94},{"className":93},[],[95],{"type":56,"value":96},"start_process",{"type":56,"value":98}," calls open separate sessions and do ",{"type":50,"tag":75,"props":100,"children":101},{},[102],{"type":56,"value":103},"not",{"type":56,"value":105}," share shell state with each other; persistence is inside one session, not across them.)",{"type":50,"tag":59,"props":107,"children":108},{},[109,114],{"type":50,"tag":75,"props":110,"children":111},{},[112],{"type":56,"value":113},"Long-running processes.",{"type":56,"value":115}," Start a dev server, watcher, build, training run, or test suite in the background and keep working. The MCP returns a process handle the agent can tail, interact with, or terminate across many turns. Long-running commands don't need to block the workflow waiting for a foreground command to exit.",{"type":50,"tag":59,"props":117,"children":118},{},[119,124],{"type":50,"tag":75,"props":120,"children":121},{},[122],{"type":56,"value":123},"Filesystem reach beyond the IDE workspace.",{"type":56,"value":125}," Read, write, move, list, and inspect files anywhere the user has granted scope — Downloads, Documents, project folders outside the IDE, or any other granted folders. Useful for organize-and-clean tasks, batch document work, and any \"look at the file my coworker just sent me\" request that doesn't fit inside the IDE sandbox.",{"type":50,"tag":59,"props":127,"children":128},{},[129,134,136,142,144,150,152,158,160,165],{"type":50,"tag":75,"props":130,"children":131},{},[132],{"type":56,"value":133},"Surgical edits to existing files.",{"type":56,"value":135}," The ",{"type":50,"tag":83,"props":137,"children":139},{"className":138},[],[140],{"type":56,"value":141},"edit_block",{"type":56,"value":143}," tool does exact-string find-and-replace with built-in safety: ambiguous matches fail loudly instead of silently overwriting the wrong thing, and an ",{"type":50,"tag":83,"props":145,"children":147},{"className":146},[],[148],{"type":56,"value":149},"expected_replacements",{"type":56,"value":151}," count prevents partial-match disasters. Lower data-loss risk than rewriting whole files based on the slice you happened to read — though a wrong ",{"type":50,"tag":83,"props":153,"children":155},{"className":154},[],[156],{"type":56,"value":157},"old_string",{"type":56,"value":159}," or wrong ",{"type":50,"tag":83,"props":161,"children":163},{"className":162},[],[164],{"type":56,"value":149},{"type":56,"value":166}," can still corrupt content, so review the changed content before considering the edit done.",{"type":50,"tag":59,"props":168,"children":169},{},[170,175],{"type":50,"tag":75,"props":171,"children":172},{},[173],{"type":56,"value":174},"Binary and structured files handled directly by the MCP.",{"type":56,"value":176}," Excel, DOCX, and PDF are first-class — read and modified through format-specific mechanisms rather than text-only approximations: Excel via cell-range JSON, DOCX via raw-XML edits, PDF via page-level operations on a new output file. The result is the real file in its original format, not a regenerated approximation. Images and PDFs return as viewable content for the agent.",{"type":50,"tag":59,"props":178,"children":179},{},[180,185],{"type":50,"tag":75,"props":181,"children":182},{},[183],{"type":56,"value":184},"Search at scale.",{"type":56,"value":186}," Streaming, ripgrep-backed search across whole projects or folder trees. The agent picks between filename search and in-file content search, pages through results progressively without flooding context, and runs multiple concurrent searches when the query is ambiguous.",{"type":50,"tag":59,"props":188,"children":189},{},[190,195],{"type":50,"tag":75,"props":191,"children":192},{},[193],{"type":56,"value":194},"Remote machines via SSH.",{"type":56,"value":196}," A long-lived SSH session inside a persistent shell turns the agent into a real ops tool: connect once, then tail logs, run diagnostics, deploy, or debug across many turns without reconnecting each step.",{"type":50,"tag":59,"props":198,"children":199},{},[200,205],{"type":50,"tag":75,"props":201,"children":202},{},[203],{"type":56,"value":204},"Process management.",{"type":56,"value":206}," List, inspect, tail, and kill accessible processes (subject to OS permissions). Useful for cleaning up stale dev servers from previous sessions and for diagnosing CPU \u002F memory issues.",{"type":50,"tag":65,"props":208,"children":210},{"id":209},"example-workflows",[211],{"type":56,"value":212},"Example workflows",{"type":50,"tag":59,"props":214,"children":215},{},[216,218,224],{"type":56,"value":217},"Each example names the actual tool sequence. Calls below are written in pseudocode shorthand (",{"type":50,"tag":83,"props":219,"children":221},{"className":220},[],[222],{"type":56,"value":223},"tool_name(\"arg\", flag=value)",{"type":56,"value":225},"); the real tools take object-shaped arguments. Tool descriptions and full parameter sets live in the MCP itself.",{"type":50,"tag":227,"props":228,"children":230},"h3",{"id":229},"debug-this-production-issue",[231],{"type":56,"value":232},"\"Debug this production issue\"",{"type":50,"tag":59,"props":234,"children":235},{},[236],{"type":56,"value":237},"Before running production-impacting SSH commands, explain the intended action and get user confirmation when the risk is non-trivial.",{"type":50,"tag":59,"props":239,"children":240},{},[241,247,249,255,257,263,265,271,273,279,281,286,288,294,296,302,304,310],{"type":50,"tag":83,"props":242,"children":244},{"className":243},[],[245],{"type":56,"value":246},"start_process(\"ssh user@prod.example.com\", timeout_ms=...)",{"type":56,"value":248}," opens a long-lived SSH session and returns a PID. ",{"type":50,"tag":83,"props":250,"children":252},{"className":251},[],[253],{"type":56,"value":254},"interact_with_process(pid, \"tail -f \u002Fvar\u002Flog\u002Fapp.log\\n\")",{"type":56,"value":256}," starts streaming logs. Subsequent turns: ",{"type":50,"tag":83,"props":258,"children":260},{"className":259},[],[261],{"type":56,"value":262},"read_process_output(pid, offset=-50)",{"type":56,"value":264}," to see the last 50 lines as they arrive, ",{"type":50,"tag":83,"props":266,"children":268},{"className":267},[],[269],{"type":56,"value":270},"interact_with_process(pid, \"...\")",{"type":56,"value":272}," to run diagnostic commands in the same session. ",{"type":50,"tag":83,"props":274,"children":276},{"className":275},[],[277],{"type":56,"value":278},"force_terminate(pid)",{"type":56,"value":280}," to close the session when done — for sessions opened by ",{"type":50,"tag":83,"props":282,"children":284},{"className":283},[],[285],{"type":56,"value":96},{"type":56,"value":287},", ",{"type":50,"tag":83,"props":289,"children":291},{"className":290},[],[292],{"type":56,"value":293},"force_terminate",{"type":56,"value":295}," is the correct cleanup tool; ",{"type":50,"tag":83,"props":297,"children":299},{"className":298},[],[300],{"type":56,"value":301},"kill_process",{"type":56,"value":303}," is for arbitrary OS PIDs found via ",{"type":50,"tag":83,"props":305,"children":307},{"className":306},[],[308],{"type":56,"value":309},"list_processes",{"type":56,"value":311},".",{"type":50,"tag":227,"props":313,"children":315},{"id":314},"deploy-this-to-staging",[316],{"type":56,"value":317},"\"Deploy this to staging\"",{"type":50,"tag":59,"props":319,"children":320},{},[321],{"type":56,"value":322},"Before deploys, restarts, migrations, or other environment-changing commands, summarize the action and confirm with the user unless they already explicitly asked for that exact operation.",{"type":50,"tag":59,"props":324,"children":325},{},[326,331,333,339,341,347,349,355,357,363],{"type":50,"tag":83,"props":327,"children":329},{"className":328},[],[330],{"type":56,"value":96},{"type":56,"value":332}," for the deploy command (could be a script, an SSH-piped command, or ",{"type":50,"tag":83,"props":334,"children":336},{"className":335},[],[337],{"type":56,"value":338},"kubectl",{"type":56,"value":340},"\u002F",{"type":50,"tag":83,"props":342,"children":344},{"className":343},[],[345],{"type":56,"value":346},"gh",{"type":56,"value":348}," etc.). ",{"type":50,"tag":83,"props":350,"children":352},{"className":351},[],[353],{"type":56,"value":354},"read_process_output",{"type":56,"value":356}," to track output and surface errors. If the deploy needs an interactive confirmation, ",{"type":50,"tag":83,"props":358,"children":360},{"className":359},[],[361],{"type":56,"value":362},"interact_with_process(pid, \"yes\\n\")",{"type":56,"value":364},". The session stays alive while the agent watches for completion or rollback.",{"type":50,"tag":227,"props":366,"children":368},{"id":367},"run-the-dev-server-and-iterate-on-the-api",[369],{"type":56,"value":370},"\"Run the dev server and iterate on the API\"",{"type":50,"tag":59,"props":372,"children":373},{},[374,380,382,387,389,395,397,403],{"type":50,"tag":83,"props":375,"children":377},{"className":376},[],[378],{"type":56,"value":379},"start_process(\"npm run dev\", timeout_ms=...)",{"type":56,"value":381}," keeps the server up. The agent then loops: ",{"type":50,"tag":83,"props":383,"children":385},{"className":384},[],[386],{"type":56,"value":141},{"type":56,"value":388}," on the route file, ",{"type":50,"tag":83,"props":390,"children":392},{"className":391},[],[393],{"type":56,"value":394},"read_process_output(pid, offset=-30)",{"type":56,"value":396}," to see the server's reload, ",{"type":50,"tag":83,"props":398,"children":400},{"className":399},[],[401],{"type":56,"value":402},"start_process(\"curl -s http:\u002F\u002Flocalhost:3000\u002Fapi\u002F...\")",{"type":56,"value":404}," for a one-shot test, repeat. The dev server never has to restart between code changes.",{"type":50,"tag":227,"props":406,"children":408},{"id":407},"refactor-across-this-monorepo",[409],{"type":56,"value":410},"\"Refactor across this monorepo\"",{"type":50,"tag":59,"props":412,"children":413},{},[414,420,422,428,430,436,438,444,446,451,453,459,461,466],{"type":50,"tag":83,"props":415,"children":417},{"className":416},[],[418],{"type":56,"value":419},"start_search(pattern=\"oldFunctionName\", path=repo_root, searchType=\"content\")",{"type":56,"value":421}," scopes every call site. ",{"type":50,"tag":83,"props":423,"children":425},{"className":424},[],[426],{"type":56,"value":427},"get_more_search_results(sessionId)",{"type":56,"value":429}," pages through. ",{"type":50,"tag":83,"props":431,"children":433},{"className":432},[],[434],{"type":56,"value":435},"read_multiple_files(paths=[...])",{"type":56,"value":437}," confirms ambiguous hits in context. ",{"type":50,"tag":83,"props":439,"children":441},{"className":440},[],[442],{"type":56,"value":443},"edit_block(file_path, old_string, new_string)",{"type":56,"value":445}," per site, with ",{"type":50,"tag":83,"props":447,"children":449},{"className":448},[],[450],{"type":56,"value":149},{"type":56,"value":452}," set when the same substring legitimately appears multiple times in one file. Verify by re-running ",{"type":50,"tag":83,"props":454,"children":456},{"className":455},[],[457],{"type":56,"value":458},"start_search",{"type":56,"value":460}," on the old name and paging the results with ",{"type":50,"tag":83,"props":462,"children":464},{"className":463},[],[465],{"type":56,"value":427},{"type":56,"value":467}," until the run completes — only then can you confirm zero remaining hits.",{"type":50,"tag":227,"props":469,"children":471},{"id":470},"update-the-q3-numbers-in-this-spreadsheet-and-tweak-the-summary-in-the-report",[472],{"type":56,"value":473},"\"Update the Q3 numbers in this spreadsheet and tweak the summary in the report\"",{"type":50,"tag":59,"props":475,"children":476},{},[477,483,485,491,493,499,501,507,509,518,520,525,527,532,534,540,542,548],{"type":50,"tag":83,"props":478,"children":480},{"className":479},[],[481],{"type":56,"value":482},"read_file(path=\"\u002F...\u002Fq3.xlsx\", sheet=\"Revenue\", range=\"A1:F50\")",{"type":56,"value":484}," returns the existing numbers as a JSON 2D array. ",{"type":50,"tag":83,"props":486,"children":488},{"className":487},[],[489],{"type":56,"value":490},"edit_block(file_path=\"\u002F...\u002Fq3.xlsx\", range=\"Revenue!C12:C24\", content=[[12345], ...])",{"type":56,"value":492}," updates the cells in place. For the report, DOCX editing is a two-read flow: first ",{"type":50,"tag":83,"props":494,"children":496},{"className":495},[],[497],{"type":56,"value":498},"read_file(path=\"\u002F...\u002Freport.docx\")",{"type":56,"value":500}," (offset 0) returns the document's outline (headings + paragraph text) so you can locate the summary section. Then ",{"type":50,"tag":83,"props":502,"children":504},{"className":503},[],[505],{"type":56,"value":506},"read_file(path=\"\u002F...\u002Freport.docx\", offset=N, length=...)",{"type":56,"value":508}," with ",{"type":50,"tag":75,"props":510,"children":511},{},[512],{"type":50,"tag":83,"props":513,"children":515},{"className":514},[],[516],{"type":56,"value":517},"N > 0",{"type":56,"value":519}," returns the raw underlying XML around that section — a non-zero offset is what flips the read into XML mode. Copy an XML fragment from that output as ",{"type":50,"tag":83,"props":521,"children":523},{"className":522},[],[524],{"type":56,"value":157},{"type":56,"value":526}," and call ",{"type":50,"tag":83,"props":528,"children":530},{"className":529},[],[531],{"type":56,"value":443},{"type":56,"value":533}," with the rewritten XML. The user gets back real ",{"type":50,"tag":83,"props":535,"children":537},{"className":536},[],[538],{"type":56,"value":539},".xlsx",{"type":56,"value":541}," and ",{"type":50,"tag":83,"props":543,"children":545},{"className":544},[],[546],{"type":56,"value":547},".docx",{"type":56,"value":549}," files, not regenerated approximations.",{"type":50,"tag":227,"props":551,"children":553},{"id":552},"generate-the-q3-report-as-a-pdf",[554],{"type":56,"value":555},"\"Generate the Q3 report as a PDF\"",{"type":50,"tag":59,"props":557,"children":558},{},[559,561,567,569,574],{"type":56,"value":560},"Compose markdown content (header, table, charts via embedded HTML), then call ",{"type":50,"tag":83,"props":562,"children":564},{"className":563},[],[565],{"type":56,"value":566},"write_pdf",{"type":56,"value":568}," to render it to a new PDF file. The MCP's ",{"type":50,"tag":83,"props":570,"children":572},{"className":571},[],[573],{"type":56,"value":566},{"type":56,"value":575}," tool description specifies the exact parameters and filename rules — follow that.",{"type":50,"tag":227,"props":577,"children":579},{"id":578},"insert-a-cover-page-into-this-pdf",[580],{"type":56,"value":581},"\"Insert a cover page into this PDF\"",{"type":50,"tag":59,"props":583,"children":584},{},[585,590,592,597],{"type":50,"tag":83,"props":586,"children":588},{"className":587},[],[589],{"type":56,"value":566},{"type":56,"value":591}," also supports modifying existing PDFs via an operations array (insert \u002F delete pages). Use it for existing-PDF edits that produce a new PDF — adding a cover page, removing a section, merging in content from another file. See the ",{"type":50,"tag":83,"props":593,"children":595},{"className":594},[],[596],{"type":56,"value":566},{"type":56,"value":598}," tool description for the operation shapes and parameter rules.",{"type":50,"tag":227,"props":600,"children":602},{"id":601},"analyze-this-200mb-csv",[603],{"type":56,"value":604},"\"Analyze this 200MB CSV\"",{"type":50,"tag":59,"props":606,"children":607},{},[608,614,616,622,624,630,631,637],{"type":50,"tag":83,"props":609,"children":611},{"className":610},[],[612],{"type":56,"value":613},"start_process(\"python3 -i\", timeout_ms=...)",{"type":56,"value":615}," opens a Python REPL and returns a PID. ",{"type":50,"tag":83,"props":617,"children":619},{"className":618},[],[620],{"type":56,"value":621},"interact_with_process(pid, \"import pandas as pd; df = pd.read_csv('\u002Fabs\u002Fpath.csv')\")",{"type":56,"value":623}," loads it once. Every subsequent question — ",{"type":50,"tag":83,"props":625,"children":627},{"className":626},[],[628],{"type":56,"value":629},"df.describe()",{"type":56,"value":287},{"type":50,"tag":83,"props":632,"children":634},{"className":633},[],[635],{"type":56,"value":636},"df.groupby('col').size()",{"type":56,"value":638},", plot a chart — runs in the same already-loaded REPL. Libraries don't re-import, the dataframe doesn't re-load. The MCP itself recommends this workflow for any local data-file analysis.",{"type":50,"tag":227,"props":640,"children":642},{"id":641},"run-a-quick-node-script",[643],{"type":56,"value":644},"\"Run a quick Node script\"",{"type":50,"tag":59,"props":646,"children":647},{},[648,654,656,661,663,669,671,676,678,684],{"type":50,"tag":83,"props":649,"children":651},{"className":650},[],[652],{"type":56,"value":653},"start_process(\"node:local\", timeout_ms=...)",{"type":56,"value":655}," opens a stateless Node execution mode on the MCP server itself — ES imports supported. ",{"type":50,"tag":83,"props":657,"children":659},{"className":658},[],[660],{"type":56,"value":96},{"type":56,"value":662}," opens the runner; each piece of JS is sent via ",{"type":50,"tag":83,"props":664,"children":666},{"className":665},[],[667],{"type":56,"value":668},"interact_with_process(pid, \"\u003Cyour JS here>\")",{"type":56,"value":670}," and runs independently (no shared state between calls). Good for one-shot transformations where keeping a long-lived REPL alive isn't worth it. Don't try to put code into the ",{"type":50,"tag":83,"props":672,"children":674},{"className":673},[],[675],{"type":56,"value":96},{"type":56,"value":677}," command argument — only the runner type (",{"type":50,"tag":83,"props":679,"children":681},{"className":680},[],[682],{"type":56,"value":683},"node:local",{"type":56,"value":685},") goes there.",{"type":50,"tag":227,"props":687,"children":689},{"id":688},"explain-this-codebase",[690],{"type":56,"value":691},"\"Explain this codebase\"",{"type":50,"tag":59,"props":693,"children":694},{},[695,701,703,709,711,717],{"type":50,"tag":83,"props":696,"children":698},{"className":697},[],[699],{"type":56,"value":700},"list_directory(path=repo_root, depth=3)",{"type":56,"value":702}," for shape. ",{"type":50,"tag":83,"props":704,"children":706},{"className":705},[],[707],{"type":56,"value":708},"start_search(pattern=\"export \", path=repo_root, searchType=\"content\")",{"type":56,"value":710}," to find the public surface. ",{"type":50,"tag":83,"props":712,"children":714},{"className":713},[],[715],{"type":56,"value":716},"read_multiple_files(paths=[entrypoints])",{"type":56,"value":718}," for the actual code. The agent can keep narrowing without re-asking the user where to look.",{"type":50,"tag":227,"props":720,"children":722},{"id":721},"organize-my-downloads-folder",[723],{"type":56,"value":724},"\"Organize my Downloads folder\"",{"type":50,"tag":59,"props":726,"children":727},{},[728,730,736,738,744,746,752,754,760,762,768,770,776],{"type":56,"value":729},"Resolve the path to absolute first (e.g., ",{"type":50,"tag":83,"props":731,"children":733},{"className":732},[],[734],{"type":56,"value":735},"\u002FUsers\u002F\u003Cuser>\u002FDownloads",{"type":56,"value":737},", not ",{"type":50,"tag":83,"props":739,"children":741},{"className":740},[],[742],{"type":56,"value":743},"~\u002FDownloads",{"type":56,"value":745},"). Then ",{"type":50,"tag":83,"props":747,"children":749},{"className":748},[],[750],{"type":56,"value":751},"list_directory(path=\"\u002FUsers\u002F\u003Cuser>\u002FDownloads\", depth=1)",{"type":56,"value":753}," to see what's there. ",{"type":50,"tag":83,"props":755,"children":757},{"className":756},[],[758],{"type":56,"value":759},"start_search(pattern=\"*.pdf\", path=\"\u002FUsers\u002F\u003Cuser>\u002FDownloads\", searchType=\"files\")",{"type":56,"value":761}," and similar for other types. ",{"type":50,"tag":83,"props":763,"children":765},{"className":764},[],[766],{"type":56,"value":767},"create_directory",{"type":56,"value":769}," for new folders. ",{"type":50,"tag":83,"props":771,"children":773},{"className":772},[],[774],{"type":56,"value":775},"move_file",{"type":56,"value":777}," per item. Preview the move plan before executing destructive ops.",{"type":50,"tag":227,"props":779,"children":781},{"id":780},"onboard-me-what-was-happening-last-session",[782],{"type":56,"value":783},"\"Onboard me — what was happening last session?\"",{"type":50,"tag":59,"props":785,"children":786},{},[787,793,795,801,803,809,811,816],{"type":50,"tag":83,"props":788,"children":790},{"className":789},[],[791],{"type":56,"value":792},"get_recent_tool_calls(maxResults=200)",{"type":56,"value":794}," returns recent activity with arguments and outputs. ",{"type":50,"tag":83,"props":796,"children":798},{"className":797},[],[799],{"type":56,"value":800},"list_sessions",{"type":56,"value":802}," shows still-running terminal sessions. ",{"type":50,"tag":83,"props":804,"children":806},{"className":805},[],[807],{"type":56,"value":808},"list_searches",{"type":56,"value":810}," shows in-flight searches. ",{"type":50,"tag":83,"props":812,"children":814},{"className":813},[],[815],{"type":56,"value":309},{"type":56,"value":817}," shows what's still alive. Together they reconstruct the work without asking the user to recap.",{"type":50,"tag":227,"props":819,"children":821},{"id":820},"why-isnt-the-repl-responding",[822],{"type":56,"value":823},"\"Why isn't the REPL responding?\"",{"type":50,"tag":59,"props":825,"children":826},{},[827,832,834,840,842,848,850,856],{"type":50,"tag":83,"props":828,"children":830},{"className":829},[],[831],{"type":56,"value":800},{"type":56,"value":833}," — if ",{"type":50,"tag":83,"props":835,"children":837},{"className":836},[],[838],{"type":56,"value":839},"Blocked: true",{"type":56,"value":841},", the REPL is waiting for input rather than hung. ",{"type":50,"tag":83,"props":843,"children":845},{"className":844},[],[846],{"type":56,"value":847},"read_process_output(pid, offset=-100)",{"type":56,"value":849}," to see what it last printed (often a prompt). ",{"type":50,"tag":83,"props":851,"children":853},{"className":852},[],[854],{"type":56,"value":855},"interact_with_process(pid, \"\u003Cthe input it's waiting for>\\n\")",{"type":56,"value":857}," unblocks it.",{"type":50,"tag":65,"props":859,"children":861},{"id":860},"core-tool-inventory",[862],{"type":56,"value":863},"Core tool inventory",{"type":50,"tag":59,"props":865,"children":866},{},[867],{"type":56,"value":868},"Grouped index of the tools an agent reaches for most often. Not exhaustive — the MCP exposes additional config \u002F diagnostics \u002F feedback tools beyond this list. Detailed parameters and return shapes for every tool are in the MCP's own tool descriptions.",{"type":50,"tag":870,"props":871,"children":872},"ul",{},[873,926,967,1001,1035],{"type":50,"tag":874,"props":875,"children":876},"li",{},[877,882,884,889,890,896,897,902,903,908,909,914,915,920,921],{"type":50,"tag":75,"props":878,"children":879},{},[880],{"type":56,"value":881},"Process \u002F shell:",{"type":56,"value":883}," ",{"type":50,"tag":83,"props":885,"children":887},{"className":886},[],[888],{"type":56,"value":96},{"type":56,"value":287},{"type":50,"tag":83,"props":891,"children":893},{"className":892},[],[894],{"type":56,"value":895},"interact_with_process",{"type":56,"value":287},{"type":50,"tag":83,"props":898,"children":900},{"className":899},[],[901],{"type":56,"value":354},{"type":56,"value":287},{"type":50,"tag":83,"props":904,"children":906},{"className":905},[],[907],{"type":56,"value":309},{"type":56,"value":287},{"type":50,"tag":83,"props":910,"children":912},{"className":911},[],[913],{"type":56,"value":800},{"type":56,"value":287},{"type":50,"tag":83,"props":916,"children":918},{"className":917},[],[919],{"type":56,"value":301},{"type":56,"value":287},{"type":50,"tag":83,"props":922,"children":924},{"className":923},[],[925],{"type":56,"value":293},{"type":50,"tag":874,"props":927,"children":928},{},[929,934,935,941,942,948,949,955,956,961,962],{"type":50,"tag":75,"props":930,"children":931},{},[932],{"type":56,"value":933},"Files (read\u002Fwrite):",{"type":56,"value":883},{"type":50,"tag":83,"props":936,"children":938},{"className":937},[],[939],{"type":56,"value":940},"read_file",{"type":56,"value":287},{"type":50,"tag":83,"props":943,"children":945},{"className":944},[],[946],{"type":56,"value":947},"read_multiple_files",{"type":56,"value":287},{"type":50,"tag":83,"props":950,"children":952},{"className":951},[],[953],{"type":56,"value":954},"write_file",{"type":56,"value":287},{"type":50,"tag":83,"props":957,"children":959},{"className":958},[],[960],{"type":56,"value":141},{"type":56,"value":287},{"type":50,"tag":83,"props":963,"children":965},{"className":964},[],[966],{"type":56,"value":566},{"type":50,"tag":874,"props":968,"children":969},{},[970,975,976,982,983,989,990,995,996],{"type":50,"tag":75,"props":971,"children":972},{},[973],{"type":56,"value":974},"Filesystem:",{"type":56,"value":883},{"type":50,"tag":83,"props":977,"children":979},{"className":978},[],[980],{"type":56,"value":981},"list_directory",{"type":56,"value":287},{"type":50,"tag":83,"props":984,"children":986},{"className":985},[],[987],{"type":56,"value":988},"get_file_info",{"type":56,"value":287},{"type":50,"tag":83,"props":991,"children":993},{"className":992},[],[994],{"type":56,"value":775},{"type":56,"value":287},{"type":50,"tag":83,"props":997,"children":999},{"className":998},[],[1000],{"type":56,"value":767},{"type":50,"tag":874,"props":1002,"children":1003},{},[1004,1009,1010,1015,1016,1022,1023,1028,1029],{"type":50,"tag":75,"props":1005,"children":1006},{},[1007],{"type":56,"value":1008},"Search:",{"type":56,"value":883},{"type":50,"tag":83,"props":1011,"children":1013},{"className":1012},[],[1014],{"type":56,"value":458},{"type":56,"value":287},{"type":50,"tag":83,"props":1017,"children":1019},{"className":1018},[],[1020],{"type":56,"value":1021},"get_more_search_results",{"type":56,"value":287},{"type":50,"tag":83,"props":1024,"children":1026},{"className":1025},[],[1027],{"type":56,"value":808},{"type":56,"value":287},{"type":50,"tag":83,"props":1030,"children":1032},{"className":1031},[],[1033],{"type":56,"value":1034},"stop_search",{"type":50,"tag":874,"props":1036,"children":1037},{},[1038,1043,1044,1050,1051],{"type":50,"tag":75,"props":1039,"children":1040},{},[1041],{"type":56,"value":1042},"Diagnostics \u002F config:",{"type":56,"value":883},{"type":50,"tag":83,"props":1045,"children":1047},{"className":1046},[],[1048],{"type":56,"value":1049},"get_recent_tool_calls",{"type":56,"value":287},{"type":50,"tag":83,"props":1052,"children":1054},{"className":1053},[],[1055],{"type":56,"value":1056},"get_config",{"type":50,"tag":65,"props":1058,"children":1060},{"id":1059},"conventions",[1061],{"type":56,"value":1062},"Conventions",{"type":50,"tag":59,"props":1064,"children":1065},{},[1066,1071,1073,1079],{"type":50,"tag":75,"props":1067,"children":1068},{},[1069],{"type":56,"value":1070},"Prefer absolute paths.",{"type":56,"value":1072}," Relative paths may fail depending on the working directory, and tilde paths (",{"type":50,"tag":83,"props":1074,"children":1076},{"className":1075},[],[1077],{"type":56,"value":1078},"~\u002F...",{"type":56,"value":1080},") may not expand in all contexts. Absolute paths are the most reliable; pass them whenever you can.",{"type":50,"tag":59,"props":1082,"children":1083},{},[1084,1089,1091,1097,1099,1105,1107,1112,1114,1119,1121,1126],{"type":50,"tag":75,"props":1085,"children":1086},{},[1087],{"type":56,"value":1088},"Allowed-directory scope.",{"type":56,"value":1090}," File operations only work inside the user's configured ",{"type":50,"tag":83,"props":1092,"children":1094},{"className":1093},[],[1095],{"type":56,"value":1096},"allowedDirectories",{"type":56,"value":1098},". Expect ",{"type":50,"tag":83,"props":1100,"children":1102},{"className":1101},[],[1103],{"type":56,"value":1104},"[DENIED]",{"type":56,"value":1106}," markers in ",{"type":50,"tag":83,"props":1108,"children":1110},{"className":1109},[],[1111],{"type":56,"value":981},{"type":56,"value":1113}," output and rejections from ",{"type":50,"tag":83,"props":1115,"children":1117},{"className":1116},[],[1118],{"type":56,"value":940},{"type":56,"value":1120}," \u002F ",{"type":50,"tag":83,"props":1122,"children":1124},{"className":1123},[],[1125],{"type":56,"value":954},{"type":56,"value":1127}," when the path is out of scope. Surface the rejected path to the user — don't retry.",{"type":50,"tag":59,"props":1129,"children":1130},{},[1131,1136,1138,1144,1146,1152,1154,1160,1162,1168,1170,1176,1178,1184,1186,1191,1193,1199,1200,1206],{"type":50,"tag":75,"props":1132,"children":1133},{},[1134],{"type":56,"value":1135},"When running on macOS:",{"type":56,"value":1137}," default shell is zsh. Use ",{"type":50,"tag":83,"props":1139,"children":1141},{"className":1140},[],[1142],{"type":56,"value":1143},"python3",{"type":56,"value":1145}," not ",{"type":50,"tag":83,"props":1147,"children":1149},{"className":1148},[],[1150],{"type":56,"value":1151},"python",{"type":56,"value":1153},". Some GNU tools have prefixed names (",{"type":50,"tag":83,"props":1155,"children":1157},{"className":1156},[],[1158],{"type":56,"value":1159},"gsed",{"type":56,"value":1161}," for GNU sed). ",{"type":50,"tag":83,"props":1163,"children":1165},{"className":1164},[],[1166],{"type":56,"value":1167},"brew",{"type":56,"value":1169}," is the typical package manager. ",{"type":50,"tag":83,"props":1171,"children":1173},{"className":1172},[],[1174],{"type":56,"value":1175},"open",{"type":56,"value":1177}," opens files \u002F apps from the terminal, ",{"type":50,"tag":83,"props":1179,"children":1181},{"className":1180},[],[1182],{"type":56,"value":1183},"mdfind",{"type":56,"value":1185}," is the fastest path to exact-filename search via Spotlight. Detect the host platform via ",{"type":50,"tag":83,"props":1187,"children":1189},{"className":1188},[],[1190],{"type":56,"value":1056},{"type":56,"value":1192}," (or by inspecting ",{"type":50,"tag":83,"props":1194,"children":1196},{"className":1195},[],[1197],{"type":56,"value":1198},"process.platform",{"type":56,"value":1120},{"type":50,"tag":83,"props":1201,"children":1203},{"className":1202},[],[1204],{"type":56,"value":1205},"uname",{"type":56,"value":1207}," from a shell) before assuming any of the above — Windows and Linux hosts behave differently.",{"type":50,"tag":59,"props":1209,"children":1210},{},[1211,1216,1218,1224,1225,1231],{"type":50,"tag":75,"props":1212,"children":1213},{},[1214],{"type":56,"value":1215},"Pagination.",{"type":56,"value":1217}," Long outputs (file reads, process output, search results) all support ",{"type":50,"tag":83,"props":1219,"children":1221},{"className":1220},[],[1222],{"type":56,"value":1223},"offset",{"type":56,"value":541},{"type":50,"tag":83,"props":1226,"children":1228},{"className":1227},[],[1229],{"type":56,"value":1230},"length",{"type":56,"value":1232},". Negative offsets read from the end (tail mode). Use these instead of dumping huge results into context.",{"items":1234,"total":1432},[1235,1254,1274,1293,1310,1326,1344,1356,1368,1386,1406,1419],{"slug":1236,"name":1236,"fn":1237,"description":1238,"org":1239,"tags":1240,"stars":31,"repoUrl":32,"updatedAt":1253},"amazon-location-service","integrate Amazon Location Service APIs","Integrates Amazon Location Service APIs for AWS applications. Use this skill when users want to add maps (interactive MapLibre or static images); geocode addresses to coordinates or reverse geocode coordinates to addresses; calculate routes, travel times, or service areas; find places and businesses through text search, nearby search, or autocomplete suggestions; retrieve detailed place information including hours, contacts, and addresses; monitor geographical boundaries with geofences; or track device locations. Covers authentication, SDK integration, and all Amazon Location Service capabilities.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1241,1244,1247,1250],{"name":1242,"slug":1243,"type":15},"API Development","api-development",{"name":1245,"slug":1246,"type":15},"AWS","aws",{"name":1248,"slug":1249,"type":15},"Maps","maps",{"name":1251,"slug":1252,"type":15},"Navigation","navigation","2026-07-12T08:13:53.294026",{"slug":1255,"name":1255,"fn":1256,"description":1257,"org":1258,"tags":1259,"stars":31,"repoUrl":32,"updatedAt":1273},"amplify-workflow","build full-stack apps with AWS Amplify","Build and deploy full-stack web and mobile apps with AWS Amplify Gen2 (TypeScript code-first). Covers auth (Cognito), data (AppSync\u002FDynamoDB including schema modeling, enum types, relationships, authorization rules), storage (S3), functions, APIs, and AI (Amplify AI Kit with Bedrock). Supports React, Next.js, Vue, Angular, React Native, Flutter, Swift, and Android. Always use this skill for Amplify Gen2 topics — even for questions you think you know — it contains validated, version-specific patterns that prevent common mistakes. TRIGGER when: user mentions Amplify Gen2; project has amplify\u002F directory or amplify_outputs; code imports @aws-amplify packages; user asks about defineBackend, defineAuth, defineData, defineStorage, or npx ampx. SKIP: Amplify Gen1 (amplify CLI v6), standalone SAM\u002FCDK without Amplify (use aws-serverless), direct Bedrock without Amplify AI Kit (use bedrock).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1260,1263,1264,1267,1270],{"name":1261,"slug":1262,"type":15},"Auth","auth",{"name":1245,"slug":1246,"type":15},{"name":1265,"slug":1266,"type":15},"Database","database",{"name":1268,"slug":1269,"type":15},"Frontend","frontend",{"name":1271,"slug":1272,"type":15},"TypeScript","typescript","2026-07-12T08:13:47.134012",{"slug":1275,"name":1275,"fn":1276,"description":1277,"org":1278,"tags":1279,"stars":31,"repoUrl":32,"updatedAt":1292},"analyzer","analyze queried data for trends","Analyze queried data for trends, week-over-week comparisons, distributions, funnels, cohorts, top-N lists, anomalies, sanity checks, and report-ready findings. Use after or alongside ClickHouse queries when the user wants insight rather than raw rows.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1280,1283,1286,1289],{"name":1281,"slug":1282,"type":15},"Analytics","analytics",{"name":1284,"slug":1285,"type":15},"ClickHouse","clickhouse",{"name":1287,"slug":1288,"type":15},"Data Analysis","data-analysis",{"name":1290,"slug":1291,"type":15},"Statistics","statistics","2026-07-12T08:14:05.606036",{"slug":1294,"name":1294,"fn":1295,"description":1296,"org":1297,"tags":1298,"stars":31,"repoUrl":32,"updatedAt":1309},"artifact-management","manage and organize analysis artifacts","Save, organize, and describe reusable analysis artifacts such as SQL, result snapshots, CSV exports, summaries, caveats, plots, and report-ready files. Use when users ask to save, export, share, cite, reproduce, or organize data-analysis outputs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1299,1300,1303,1306],{"name":1287,"slug":1288,"type":15},{"name":1301,"slug":1302,"type":15},"Productivity","productivity",{"name":1304,"slug":1305,"type":15},"Reporting","reporting",{"name":1307,"slug":1308,"type":15},"SQL","sql","2026-07-12T08:14:09.265555",{"slug":1311,"name":1311,"fn":1312,"description":1313,"org":1314,"tags":1315,"stars":31,"repoUrl":32,"updatedAt":1325},"attorney-assist","connect with attorneys for legal consultation","Connects the user with a LegalZoom attorney for legal consultation. Use when a user asks about attorneys, lawyers, or legal help, or when contract review reveals high risks or low-confidence findings.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1316,1319,1322],{"name":1317,"slug":1318,"type":15},"Contracts","contracts",{"name":1320,"slug":1321,"type":15},"Legal","legal",{"name":1323,"slug":1324,"type":15},"Risk Assessment","risk-assessment","2026-07-12T08:13:45.878361",{"slug":1327,"name":1327,"fn":1328,"description":1329,"org":1330,"tags":1331,"stars":31,"repoUrl":32,"updatedAt":1343},"building-pydantic-ai-agents","build AI agents with Pydantic AI","Build AI agents with Pydantic AI — tools, capabilities (including on-demand loading), structured output, streaming, testing, and multi-agent patterns. Use when the user mentions Pydantic AI, imports pydantic_ai, or asks to build an AI agent, add tools\u002Fcapabilities, defer capability loading, stream output, define agents from YAML, or test agent behavior.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1332,1335,1338,1341],{"name":1333,"slug":1334,"type":15},"Agents","agents",{"name":1336,"slug":1337,"type":15},"LLM","llm",{"name":1339,"slug":1340,"type":15},"Multi-Agent","multi-agent",{"name":1342,"slug":1151,"type":15},"Python","2026-07-12T08:14:01.893781",{"slug":1285,"name":1285,"fn":1345,"description":1346,"org":1347,"tags":1348,"stars":31,"repoUrl":32,"updatedAt":1355},"query ClickHouse databases via CLI","Connect to and query ClickHouse (a local server or a ClickHouse Cloud service) from the terminal using the official clickhousectl CLI, including the browser OAuth login flow. Use when the user wants to run SQL against ClickHouse, explore schemas and tables, inspect Cloud services, or authenticate clickhousectl. For building a local dev environment or deploying to Cloud, defer to the official ClickHouse skills (see Scope).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1349,1350,1353,1354],{"name":1281,"slug":1282,"type":15},{"name":1351,"slug":1352,"type":15},"CLI","cli",{"name":1284,"slug":1285,"type":15},{"name":1265,"slug":1266,"type":15},"2026-07-12T08:14:06.829692",{"slug":1357,"name":1357,"fn":1358,"description":1359,"org":1360,"tags":1361,"stars":31,"repoUrl":32,"updatedAt":1367},"cline-session-history","search and browse Cline session history","Search and browse Cline session history. Use when the user asks to find, list, inspect, or export Cline sessions by time period, prompt content, status, model, provider, source, mode, workspace, or other criteria. Also use when the user wants to read a session transcript or export sessions to a directory. Trigger phrases include \"find my session\", \"search my session history\", \"show me past sessions\", \"what was that session where\", \"find the session that started with\", and any mention of past Cline conversations.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1362,1363,1366],{"name":1333,"slug":1334,"type":15},{"name":1364,"slug":1365,"type":15},"History","history",{"name":1301,"slug":1302,"type":15},"2026-07-19T06:03:13.945151",{"slug":1369,"name":1369,"fn":1370,"description":1371,"org":1372,"tags":1373,"stars":31,"repoUrl":32,"updatedAt":1385},"convex-design","build reactive backends with Convex","Design and build reactive, type-safe, production-grade backends on Convex. Covers schema, queries\u002Fmutations\u002Factions, indexes, auth, file storage, scheduling, real-time multiplayer, mobile backends, and LLM\u002Fagent workflows on Convex's one-platform stack.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1374,1375,1378,1379,1382],{"name":1261,"slug":1262,"type":15},{"name":1376,"slug":1377,"type":15},"Backend","backend",{"name":1265,"slug":1266,"type":15},{"name":1380,"slug":1381,"type":15},"Real-time","real-time",{"name":1383,"slug":1384,"type":15},"Storage","storage","2026-07-12T08:13:37.101253",{"slug":1387,"name":1387,"fn":1388,"description":1389,"org":1390,"tags":1391,"stars":31,"repoUrl":32,"updatedAt":1405},"cosmosdb-best-practices","optimize Azure Cosmos DB performance","Azure Cosmos DB performance optimization and best practices guidelines for NoSQL,\npartitioning, queries, SDK usage, and vector search. Use when writing, reviewing,\nor refactoring code that interacts with Azure Cosmos DB, designing data models,\noptimizing queries, or implementing high-performance database operations.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1392,1395,1398,1399,1402],{"name":1393,"slug":1394,"type":15},"Azure","azure",{"name":1396,"slug":1397,"type":15},"Cosmos DB","cosmos-db",{"name":1265,"slug":1266,"type":15},{"name":1400,"slug":1401,"type":15},"NoSQL","nosql",{"name":1403,"slug":1404,"type":15},"Performance","performance","2026-07-12T08:13:54.531719",{"slug":1407,"name":1407,"fn":1408,"description":1409,"org":1410,"tags":1411,"stars":31,"repoUrl":32,"updatedAt":1418},"data-analyst","analyze ClickHouse analytics data","Act as an interactive data analyst for ClickHouse-backed analytics. Use when the user asks questions about internal data, metrics, dashboards, telemetry, active users, revenue, funnels, trends, distributions, or wants an analyst-style conversation, ad hoc SQL, charts, or a data export against ClickHouse (local or ClickHouse Cloud).",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1412,1413,1414,1417],{"name":1281,"slug":1282,"type":15},{"name":1284,"slug":1285,"type":15},{"name":1415,"slug":1416,"type":15},"Dashboards","dashboards",{"name":1287,"slug":1288,"type":15},"2026-07-12T08:13:31.975246",{"slug":1420,"name":1420,"fn":1421,"description":1422,"org":1423,"tags":1424,"stars":31,"repoUrl":32,"updatedAt":1431},"dataproc-skills","manage Dataproc clusters and jobs","Skills to interact with your Dataproc clusters and jobs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1425,1428],{"name":1426,"slug":1427,"type":15},"Data Engineering","data-engineering",{"name":1429,"slug":1430,"type":15},"Operations","operations","2026-07-12T08:13:42.179275",45,{"items":1434,"total":1484},[1435,1442,1450,1457,1464,1470,1477],{"slug":1236,"name":1236,"fn":1237,"description":1238,"org":1436,"tags":1437,"stars":31,"repoUrl":32,"updatedAt":1253},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1438,1439,1440,1441],{"name":1242,"slug":1243,"type":15},{"name":1245,"slug":1246,"type":15},{"name":1248,"slug":1249,"type":15},{"name":1251,"slug":1252,"type":15},{"slug":1255,"name":1255,"fn":1256,"description":1257,"org":1443,"tags":1444,"stars":31,"repoUrl":32,"updatedAt":1273},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1445,1446,1447,1448,1449],{"name":1261,"slug":1262,"type":15},{"name":1245,"slug":1246,"type":15},{"name":1265,"slug":1266,"type":15},{"name":1268,"slug":1269,"type":15},{"name":1271,"slug":1272,"type":15},{"slug":1275,"name":1275,"fn":1276,"description":1277,"org":1451,"tags":1452,"stars":31,"repoUrl":32,"updatedAt":1292},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1453,1454,1455,1456],{"name":1281,"slug":1282,"type":15},{"name":1284,"slug":1285,"type":15},{"name":1287,"slug":1288,"type":15},{"name":1290,"slug":1291,"type":15},{"slug":1294,"name":1294,"fn":1295,"description":1296,"org":1458,"tags":1459,"stars":31,"repoUrl":32,"updatedAt":1309},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1460,1461,1462,1463],{"name":1287,"slug":1288,"type":15},{"name":1301,"slug":1302,"type":15},{"name":1304,"slug":1305,"type":15},{"name":1307,"slug":1308,"type":15},{"slug":1311,"name":1311,"fn":1312,"description":1313,"org":1465,"tags":1466,"stars":31,"repoUrl":32,"updatedAt":1325},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1467,1468,1469],{"name":1317,"slug":1318,"type":15},{"name":1320,"slug":1321,"type":15},{"name":1323,"slug":1324,"type":15},{"slug":1327,"name":1327,"fn":1328,"description":1329,"org":1471,"tags":1472,"stars":31,"repoUrl":32,"updatedAt":1343},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1473,1474,1475,1476],{"name":1333,"slug":1334,"type":15},{"name":1336,"slug":1337,"type":15},{"name":1339,"slug":1340,"type":15},{"name":1342,"slug":1151,"type":15},{"slug":1285,"name":1285,"fn":1345,"description":1346,"org":1478,"tags":1479,"stars":31,"repoUrl":32,"updatedAt":1355},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1480,1481,1482,1483],{"name":1281,"slug":1282,"type":15},{"name":1351,"slug":1352,"type":15},{"name":1284,"slug":1285,"type":15},{"name":1265,"slug":1266,"type":15},43]