[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-microsoft-systems-design-review-methodology":3,"mdc--8rq7ld-key":33,"related-repo-microsoft-systems-design-review-methodology":1286,"related-org-microsoft-systems-design-review-methodology":1362},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":22,"repoUrl":23,"updatedAt":24,"license":25,"forks":26,"topics":27,"repo":28,"sourceUrl":31,"mdContent":32},"systems-design-review-methodology","perform systems design reviews","Use when the \u002Fsystems-design-review mode is active. 7-step design review methodology -- understand the design, classify the system, evaluate against codebase, adversarial analysis, tradeoff validation, synthesis, and action items. Governs conversation flow, delegation patterns, and user validation checkpoints.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"microsoft","Microsoft","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fmicrosoft.png",[12,16,19],{"name":13,"slug":14,"type":15},"Architecture","architecture","tag",{"name":17,"slug":18,"type":15},"System Design","system-design",{"name":20,"slug":21,"type":15},"Engineering","engineering",0,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-bundle-systems-design","2026-07-03T16:31:58.725837",null,1,[],{"repoUrl":23,"stars":22,"forks":26,"topics":29,"description":30},[],"Systems-design capabilities for Amplifier sessions.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Famplifier-bundle-systems-design\u002Ftree\u002FHEAD\u002Fskills\u002Fsystems-design-review-methodology","---\nname: systems-design-review-methodology\ndescription: \"Use when the \u002Fsystems-design-review mode is active. 7-step design review methodology -- understand the design, classify the system, evaluate against codebase, adversarial analysis, tradeoff validation, synthesis, and action items. Governs conversation flow, delegation patterns, and user validation checkpoints.\"\n---\n\n# Design Review Methodology\n\nCompanion skill for the `\u002Fsystems-design-review` mode. The mode gates tools; this skill governs behavior.\n\n**For automated staging with approval gates**, use the `systems-design-review` recipe instead of this manual flow:\n```\nrecipes(operation=\"execute\", recipe_path=\"@systems-design:recipes\u002Fsystems-design-review.yaml\", context={\"target_path\": \"\u003Cpath>\"})\n```\nThe recipe automates reconnaissance, classification, multi-perspective analysis, and report generation with human checkpoints between stages.\n\n## Core Rule\n\n**You handle the CONVERSATION. Agents handle the ANALYSIS.**\n\nYou guide the user through the review, synthesize findings, and facilitate decisions. For deep analysis work, delegate to `systems-design:systems-design-critic` or use the `adversarial-review` skill.\n\n## Step Flow\n\n### Step 1: Understand the Design\n\nBefore any critique, reconstruct the designer's reasoning:\n\n1. Read the design document or referenced files thoroughly.\n2. Identify the stated goals, constraints, and tradeoffs.\n3. If a codebase exists, survey the relevant parts.\n4. Note what the design explicitly addresses and what it is silent on.\n\n**Apply the Comprehending Existing lens:**\n- Why was this designed this way? What tradeoffs did the original designer make?\n- What does the choice of this approach over alternatives tell us about the constraints the designer was working under?\n- What is the mental model? Is it human-facing or system-facing?\n- What problems does the current design solve that aren't obvious at first glance?\n\nUnderstanding original intent prevents reviews that recommend \"fixing\" things the designer already considered and deliberately chose.\n\nAsk: **\"Is this the complete design, or is there additional context I should know about?\"**\n\n### Step 2: Classify the System\n\n**This step is mandatory. Do not skip it. Do not proceed to analysis without completing it.**\n\nBased on what you learned in Step 1, classify the system under review. Produce a brief taxonomy:\n\n**System types** -- which of these apply? List ALL that match, not just the primary one:\n\n| Type | Skill | Applies when... |\n|------|-------|-----------------| \n| Web service \u002F API | `system-type-web-service` | HTTP endpoints, REST\u002FGraphQL, request-response |\n| Event-driven | `system-type-event-driven` | Message queues, event logs, pub\u002Fsub, hooks, reactive patterns |\n| Data pipeline | `system-type-data-pipeline` | Batch\u002Fstreaming processing, ETL, DAG scheduling |\n| Workflow orchestration | `system-type-workflow-orchestration` | Multi-step processes, sagas, durable execution |\n| CLI tool | `system-type-cli-tool` | Command-line interface, subcommands, plugin architecture |\n| Real-time | `system-type-real-time` | WebSockets, persistent connections, state sync |\n| Multi-tenant SaaS | `system-type-multi-tenant-saas` | Tenant isolation, shared infrastructure, billing |\n| ML serving | `system-type-ml-serving` | Model serving, feature stores, inference pipelines |\n| Distributed system | `system-type-distributed` | Consensus, replication, partitioning, multi-node coordination |\n| Enterprise integration | `system-type-enterprise-integration` | Legacy modernization, API gateways, data integration |\n| Edge \u002F offline-first | `system-type-edge-offline` | Offline operation, sync protocols, constrained resources |\n| Single-page app | `system-type-spa` | Client-side routing, state management, rendering strategies |\n| Peer-to-peer | `system-type-peer-to-peer` | P2P topologies, NAT traversal, decentralized coordination |\n| Azure-hosted | `system-type-azure` | Azure compute, identity, networking, managed services |\n\n**Design philosophies** -- which does the system claim or embody?\n\n| Philosophy | Skill | Applies when... |\n|------------|-------|-----------------| \n| Linux\u002FUnix | `design-philosophy-linux` | Mechanism vs policy, composability, small sharp tools |\n| Domain-driven | `design-philosophy-domain-driven` | Bounded contexts, ubiquitous language, aggregates |\n| Object-oriented | `design-philosophy-object-oriented` | SOLID, composition over inheritance, protocols\u002Ftraits |\n\n**After classifying, immediately load ALL matching skills:**\n```\nload_skill(skill_name=\"system-type-event-driven\")\nload_skill(skill_name=\"system-type-cli-tool\")\nload_skill(skill_name=\"design-philosophy-linux\")\n# ... every skill that matches\n```\n\nPresent the classification to the user: **\"I've classified this system as [types]. I've loaded domain skills for [list]. Does this classification match your understanding? Anything I should add?\"**\n\n**Why this matters:** Domain skills contain failure modes, anti-patterns, and evaluation criteria specific to each system type. Without them, the review produces generic findings. With them, findings are grounded in domain-specific patterns -- \"this violates event-driven reliability patterns\" rather than \"this might have a timeout issue.\"\n\n### Step 3: Evaluate Against the Codebase\n\nIf reviewing an existing system or a design that modifies one:\n- Use `read_file`, `grep`, and `LSP` to examine the actual code\n- Compare the design's claims against reality\n- Identify gaps between the design document and the implementation\n\n**Implementation viability check:** If the design proposes changes to an existing system, trace the impact exhaustively:\n- Use LSP `findReferences` and `incomingCalls` to find every usage of components being changed\n- For each usage, validate the new design can replace the old seamlessly\n- Flag any callsite where the replacement is non-trivial or ambiguous\n\nThis is not optional exploration -- it is the mechanism that prevents \"it worked in theory but broke 12 callsites\" surprises.\n\n### Step 4: Adversarial Analysis\n\nDelegate the deep adversarial review rather than doing it inline:\n\n**Option A -- Delegate to the systems-design-critic agent:**\n```\ndelegate(\n  agent=\"systems-design:systems-design-critic\",\n  instruction=\"Review this design: [design content or file path]. System context: [relevant codebase findings from Step 3]. System classification: [types from Step 2]\",\n  context_depth=\"recent\",\n  context_scope=\"agents\"\n)\n```\n\n**Option B -- Use the adversarial-review skill (spawns 6 parallel agents):**\n```\nload_skill(skill_name=\"adversarial-review\")\n```\n\nChoose Option A when you have accumulated codebase context to pass along. Choose Option B for a quick parallel review when the design is self-contained.\n\n**Pass the system classification to the agent.** The critic should know what type of system it is reviewing so it can apply domain-specific failure modes rather than generic concerns.\n\n### Step 5: Tradeoff Validation\n\nAfter the adversarial analysis returns, load the tradeoff-analysis skill if the review reveals unclear tradeoffs:\n```\nload_skill(skill_name=\"tradeoff-analysis\")\n```\n\nVerify:\n- Are the stated tradeoffs actually what the design achieves?\n- Are there unstated tradeoffs the designers may not have recognized?\n- Does the design optimize for what the team actually needs?\n\n**Apply the Simplicity and DX lenses:**\n- Isn't there a simpler way to accomplish what this design does? What concept have we uncovered that we didn't know before?\n- Are we mixing two concerns here? Aren't these actually separate responsibilities?\n- How many concepts must a developer hold in their head to use this correctly? Is the naming consistent and clear?\n- Does this diverge from ecosystem conventions? If so, is the divergence justified?\n\n### Step 6: Synthesize and Recommend\n\nPresent findings to the user.\n\n**Before presenting, categorize each finding:**\n- **Architectural** -- boundaries, data flow, failure modes, coupling, scalability, consistency model, state ownership. These belong in a systems-design review.\n- **Implementation** -- code quality, specific function issues, API ergonomics, type safety, naming. These are useful but secondary.\n\nIf more than 30% of findings are implementation-level, you have drifted from the systems-design scope. Reframe implementation findings in architectural terms or separate them into a distinct section.\n\n**Check design integrity before finalizing:**\n- Is this design consistent with previous design decisions in the system? If not, is the inconsistency explicitly reconciled?\n- Who owns this component? Does it really belong here, or are we changing the responsibility of something established?\n- Does this introduce a second way to do something the system already does?\n\n**Assessment Summary:**\n- Overall health: [strong \u002F has concerns \u002F needs significant revision]\n- Critical risks: [list]\n- Tradeoff alignment: [does the design achieve what it claims?]\n\n**Recommendation:** One of:\n- **Proceed** -- risks are acceptable, design is sound\n- **Proceed with modifications** -- [specific changes needed]\n- **Reconsider** -- [fundamental issue that requires rethinking]\n\nAsk: **\"Do these findings match your intuition? Are there aspects of the design I should look at more closely?\"**\n\n### Step 7: Capture Action Items\n\nIf the user wants to proceed with modifications, produce a clear action list:\n1. What to change (specific, not vague)\n2. Why (linked to a specific risk or finding)\n3. Priority (must-do vs should-do vs nice-to-do)\n\n## Handling Follow-Up Questions After Agent Work\n\nWhen agents have already completed analysis and the user asks a follow-up question, do NOT re-delegate from scratch. Prior agent work is expensive -- treat it as an asset, not a disposable draft.\n\n**Decision rule:**\n\n1. **Can I answer from prior agent findings?** If the architect or critic already analyzed this system, synthesize their findings into a response. This is conversation work, not analysis work -- it belongs to you.\n2. **Does the follow-up require genuinely new analysis?** New scope, new data, or a different analytical frame that prior work didn't cover. Only then re-delegate.\n3. **If re-delegating, build on prior work.** Resume the existing agent session (`session_id`) when possible. If spawning fresh, include a summary of prior findings in the instruction so the agent extends rather than repeats.\n\n**The test:** If the new delegation instruction overlaps >50% with a prior one, you are re-doing work. Synthesize instead.\n\n**Anti-pattern:** User asks about higher-level architectural decisions. You feel uncertain about your own synthesis, so you re-spawn the architect and critic from scratch. Both agents produce findings that overlap heavily with their first pass. Hours of computation wasted, and the user watches agents return similar results twice.\n\n**Correct pattern:** User asks about architectural decisions. You check: the architect already assessed boundaries, coupling, and failure modes; the critic already reviewed from 6 perspectives. You synthesize those findings into an answer focused on the user's specific question. If the question touches something genuinely uncovered (e.g., a specific subsystem the prior analysis didn't examine), you delegate only that gap -- and reference prior findings so the agent builds on them.\n\n**The \"analysis vs conversation\" heuristic:** If answering requires reasoning about tradeoffs, failure modes, or design alternatives using information you don't have, it is ANALYSIS -- delegate it. If answering requires presenting, reframing, or connecting findings that agents already produced, it is CONVERSATION -- do it yourself. When in doubt, check whether prior agent results already contain the answer before delegating.\n",{"data":34,"body":35},{"name":4,"description":6},{"type":36,"children":37},"root",[38,47,62,81,93,98,105,113,134,140,147,152,177,185,209,214,224,230,238,243,253,594,604,695,703,712,737,747,753,758,800,810,844,849,855,860,868,877,885,894,899,909,915,920,929,934,952,960,983,989,994,1002,1025,1030,1038,1056,1064,1096,1106,1148,1157,1163,1168,1186,1192,1197,1205,1246,1256,1266,1276],{"type":39,"tag":40,"props":41,"children":43},"element","h1",{"id":42},"design-review-methodology",[44],{"type":45,"value":46},"text","Design Review Methodology",{"type":39,"tag":48,"props":49,"children":50},"p",{},[51,53,60],{"type":45,"value":52},"Companion skill for the ",{"type":39,"tag":54,"props":55,"children":57},"code",{"className":56},[],[58],{"type":45,"value":59},"\u002Fsystems-design-review",{"type":45,"value":61}," mode. The mode gates tools; this skill governs behavior.",{"type":39,"tag":48,"props":63,"children":64},{},[65,71,73,79],{"type":39,"tag":66,"props":67,"children":68},"strong",{},[69],{"type":45,"value":70},"For automated staging with approval gates",{"type":45,"value":72},", use the ",{"type":39,"tag":54,"props":74,"children":76},{"className":75},[],[77],{"type":45,"value":78},"systems-design-review",{"type":45,"value":80}," recipe instead of this manual flow:",{"type":39,"tag":82,"props":83,"children":87},"pre",{"className":84,"code":86,"language":45},[85],"language-text","recipes(operation=\"execute\", recipe_path=\"@systems-design:recipes\u002Fsystems-design-review.yaml\", context={\"target_path\": \"\u003Cpath>\"})\n",[88],{"type":39,"tag":54,"props":89,"children":91},{"__ignoreMap":90},"",[92],{"type":45,"value":86},{"type":39,"tag":48,"props":94,"children":95},{},[96],{"type":45,"value":97},"The recipe automates reconnaissance, classification, multi-perspective analysis, and report generation with human checkpoints between stages.",{"type":39,"tag":99,"props":100,"children":102},"h2",{"id":101},"core-rule",[103],{"type":45,"value":104},"Core Rule",{"type":39,"tag":48,"props":106,"children":107},{},[108],{"type":39,"tag":66,"props":109,"children":110},{},[111],{"type":45,"value":112},"You handle the CONVERSATION. Agents handle the ANALYSIS.",{"type":39,"tag":48,"props":114,"children":115},{},[116,118,124,126,132],{"type":45,"value":117},"You guide the user through the review, synthesize findings, and facilitate decisions. For deep analysis work, delegate to ",{"type":39,"tag":54,"props":119,"children":121},{"className":120},[],[122],{"type":45,"value":123},"systems-design:systems-design-critic",{"type":45,"value":125}," or use the ",{"type":39,"tag":54,"props":127,"children":129},{"className":128},[],[130],{"type":45,"value":131},"adversarial-review",{"type":45,"value":133}," skill.",{"type":39,"tag":99,"props":135,"children":137},{"id":136},"step-flow",[138],{"type":45,"value":139},"Step Flow",{"type":39,"tag":141,"props":142,"children":144},"h3",{"id":143},"step-1-understand-the-design",[145],{"type":45,"value":146},"Step 1: Understand the Design",{"type":39,"tag":48,"props":148,"children":149},{},[150],{"type":45,"value":151},"Before any critique, reconstruct the designer's reasoning:",{"type":39,"tag":153,"props":154,"children":155},"ol",{},[156,162,167,172],{"type":39,"tag":157,"props":158,"children":159},"li",{},[160],{"type":45,"value":161},"Read the design document or referenced files thoroughly.",{"type":39,"tag":157,"props":163,"children":164},{},[165],{"type":45,"value":166},"Identify the stated goals, constraints, and tradeoffs.",{"type":39,"tag":157,"props":168,"children":169},{},[170],{"type":45,"value":171},"If a codebase exists, survey the relevant parts.",{"type":39,"tag":157,"props":173,"children":174},{},[175],{"type":45,"value":176},"Note what the design explicitly addresses and what it is silent on.",{"type":39,"tag":48,"props":178,"children":179},{},[180],{"type":39,"tag":66,"props":181,"children":182},{},[183],{"type":45,"value":184},"Apply the Comprehending Existing lens:",{"type":39,"tag":186,"props":187,"children":188},"ul",{},[189,194,199,204],{"type":39,"tag":157,"props":190,"children":191},{},[192],{"type":45,"value":193},"Why was this designed this way? What tradeoffs did the original designer make?",{"type":39,"tag":157,"props":195,"children":196},{},[197],{"type":45,"value":198},"What does the choice of this approach over alternatives tell us about the constraints the designer was working under?",{"type":39,"tag":157,"props":200,"children":201},{},[202],{"type":45,"value":203},"What is the mental model? Is it human-facing or system-facing?",{"type":39,"tag":157,"props":205,"children":206},{},[207],{"type":45,"value":208},"What problems does the current design solve that aren't obvious at first glance?",{"type":39,"tag":48,"props":210,"children":211},{},[212],{"type":45,"value":213},"Understanding original intent prevents reviews that recommend \"fixing\" things the designer already considered and deliberately chose.",{"type":39,"tag":48,"props":215,"children":216},{},[217,219],{"type":45,"value":218},"Ask: ",{"type":39,"tag":66,"props":220,"children":221},{},[222],{"type":45,"value":223},"\"Is this the complete design, or is there additional context I should know about?\"",{"type":39,"tag":141,"props":225,"children":227},{"id":226},"step-2-classify-the-system",[228],{"type":45,"value":229},"Step 2: Classify the System",{"type":39,"tag":48,"props":231,"children":232},{},[233],{"type":39,"tag":66,"props":234,"children":235},{},[236],{"type":45,"value":237},"This step is mandatory. Do not skip it. Do not proceed to analysis without completing it.",{"type":39,"tag":48,"props":239,"children":240},{},[241],{"type":45,"value":242},"Based on what you learned in Step 1, classify the system under review. Produce a brief taxonomy:",{"type":39,"tag":48,"props":244,"children":245},{},[246,251],{"type":39,"tag":66,"props":247,"children":248},{},[249],{"type":45,"value":250},"System types",{"type":45,"value":252}," -- which of these apply? List ALL that match, not just the primary one:",{"type":39,"tag":254,"props":255,"children":256},"table",{},[257,281],{"type":39,"tag":258,"props":259,"children":260},"thead",{},[261],{"type":39,"tag":262,"props":263,"children":264},"tr",{},[265,271,276],{"type":39,"tag":266,"props":267,"children":268},"th",{},[269],{"type":45,"value":270},"Type",{"type":39,"tag":266,"props":272,"children":273},{},[274],{"type":45,"value":275},"Skill",{"type":39,"tag":266,"props":277,"children":278},{},[279],{"type":45,"value":280},"Applies when...",{"type":39,"tag":282,"props":283,"children":284},"tbody",{},[285,308,330,352,374,396,418,440,462,484,506,528,550,572],{"type":39,"tag":262,"props":286,"children":287},{},[288,294,303],{"type":39,"tag":289,"props":290,"children":291},"td",{},[292],{"type":45,"value":293},"Web service \u002F API",{"type":39,"tag":289,"props":295,"children":296},{},[297],{"type":39,"tag":54,"props":298,"children":300},{"className":299},[],[301],{"type":45,"value":302},"system-type-web-service",{"type":39,"tag":289,"props":304,"children":305},{},[306],{"type":45,"value":307},"HTTP endpoints, REST\u002FGraphQL, request-response",{"type":39,"tag":262,"props":309,"children":310},{},[311,316,325],{"type":39,"tag":289,"props":312,"children":313},{},[314],{"type":45,"value":315},"Event-driven",{"type":39,"tag":289,"props":317,"children":318},{},[319],{"type":39,"tag":54,"props":320,"children":322},{"className":321},[],[323],{"type":45,"value":324},"system-type-event-driven",{"type":39,"tag":289,"props":326,"children":327},{},[328],{"type":45,"value":329},"Message queues, event logs, pub\u002Fsub, hooks, reactive patterns",{"type":39,"tag":262,"props":331,"children":332},{},[333,338,347],{"type":39,"tag":289,"props":334,"children":335},{},[336],{"type":45,"value":337},"Data pipeline",{"type":39,"tag":289,"props":339,"children":340},{},[341],{"type":39,"tag":54,"props":342,"children":344},{"className":343},[],[345],{"type":45,"value":346},"system-type-data-pipeline",{"type":39,"tag":289,"props":348,"children":349},{},[350],{"type":45,"value":351},"Batch\u002Fstreaming processing, ETL, DAG scheduling",{"type":39,"tag":262,"props":353,"children":354},{},[355,360,369],{"type":39,"tag":289,"props":356,"children":357},{},[358],{"type":45,"value":359},"Workflow orchestration",{"type":39,"tag":289,"props":361,"children":362},{},[363],{"type":39,"tag":54,"props":364,"children":366},{"className":365},[],[367],{"type":45,"value":368},"system-type-workflow-orchestration",{"type":39,"tag":289,"props":370,"children":371},{},[372],{"type":45,"value":373},"Multi-step processes, sagas, durable execution",{"type":39,"tag":262,"props":375,"children":376},{},[377,382,391],{"type":39,"tag":289,"props":378,"children":379},{},[380],{"type":45,"value":381},"CLI tool",{"type":39,"tag":289,"props":383,"children":384},{},[385],{"type":39,"tag":54,"props":386,"children":388},{"className":387},[],[389],{"type":45,"value":390},"system-type-cli-tool",{"type":39,"tag":289,"props":392,"children":393},{},[394],{"type":45,"value":395},"Command-line interface, subcommands, plugin architecture",{"type":39,"tag":262,"props":397,"children":398},{},[399,404,413],{"type":39,"tag":289,"props":400,"children":401},{},[402],{"type":45,"value":403},"Real-time",{"type":39,"tag":289,"props":405,"children":406},{},[407],{"type":39,"tag":54,"props":408,"children":410},{"className":409},[],[411],{"type":45,"value":412},"system-type-real-time",{"type":39,"tag":289,"props":414,"children":415},{},[416],{"type":45,"value":417},"WebSockets, persistent connections, state sync",{"type":39,"tag":262,"props":419,"children":420},{},[421,426,435],{"type":39,"tag":289,"props":422,"children":423},{},[424],{"type":45,"value":425},"Multi-tenant SaaS",{"type":39,"tag":289,"props":427,"children":428},{},[429],{"type":39,"tag":54,"props":430,"children":432},{"className":431},[],[433],{"type":45,"value":434},"system-type-multi-tenant-saas",{"type":39,"tag":289,"props":436,"children":437},{},[438],{"type":45,"value":439},"Tenant isolation, shared infrastructure, billing",{"type":39,"tag":262,"props":441,"children":442},{},[443,448,457],{"type":39,"tag":289,"props":444,"children":445},{},[446],{"type":45,"value":447},"ML serving",{"type":39,"tag":289,"props":449,"children":450},{},[451],{"type":39,"tag":54,"props":452,"children":454},{"className":453},[],[455],{"type":45,"value":456},"system-type-ml-serving",{"type":39,"tag":289,"props":458,"children":459},{},[460],{"type":45,"value":461},"Model serving, feature stores, inference pipelines",{"type":39,"tag":262,"props":463,"children":464},{},[465,470,479],{"type":39,"tag":289,"props":466,"children":467},{},[468],{"type":45,"value":469},"Distributed system",{"type":39,"tag":289,"props":471,"children":472},{},[473],{"type":39,"tag":54,"props":474,"children":476},{"className":475},[],[477],{"type":45,"value":478},"system-type-distributed",{"type":39,"tag":289,"props":480,"children":481},{},[482],{"type":45,"value":483},"Consensus, replication, partitioning, multi-node coordination",{"type":39,"tag":262,"props":485,"children":486},{},[487,492,501],{"type":39,"tag":289,"props":488,"children":489},{},[490],{"type":45,"value":491},"Enterprise integration",{"type":39,"tag":289,"props":493,"children":494},{},[495],{"type":39,"tag":54,"props":496,"children":498},{"className":497},[],[499],{"type":45,"value":500},"system-type-enterprise-integration",{"type":39,"tag":289,"props":502,"children":503},{},[504],{"type":45,"value":505},"Legacy modernization, API gateways, data integration",{"type":39,"tag":262,"props":507,"children":508},{},[509,514,523],{"type":39,"tag":289,"props":510,"children":511},{},[512],{"type":45,"value":513},"Edge \u002F offline-first",{"type":39,"tag":289,"props":515,"children":516},{},[517],{"type":39,"tag":54,"props":518,"children":520},{"className":519},[],[521],{"type":45,"value":522},"system-type-edge-offline",{"type":39,"tag":289,"props":524,"children":525},{},[526],{"type":45,"value":527},"Offline operation, sync protocols, constrained resources",{"type":39,"tag":262,"props":529,"children":530},{},[531,536,545],{"type":39,"tag":289,"props":532,"children":533},{},[534],{"type":45,"value":535},"Single-page app",{"type":39,"tag":289,"props":537,"children":538},{},[539],{"type":39,"tag":54,"props":540,"children":542},{"className":541},[],[543],{"type":45,"value":544},"system-type-spa",{"type":39,"tag":289,"props":546,"children":547},{},[548],{"type":45,"value":549},"Client-side routing, state management, rendering strategies",{"type":39,"tag":262,"props":551,"children":552},{},[553,558,567],{"type":39,"tag":289,"props":554,"children":555},{},[556],{"type":45,"value":557},"Peer-to-peer",{"type":39,"tag":289,"props":559,"children":560},{},[561],{"type":39,"tag":54,"props":562,"children":564},{"className":563},[],[565],{"type":45,"value":566},"system-type-peer-to-peer",{"type":39,"tag":289,"props":568,"children":569},{},[570],{"type":45,"value":571},"P2P topologies, NAT traversal, decentralized coordination",{"type":39,"tag":262,"props":573,"children":574},{},[575,580,589],{"type":39,"tag":289,"props":576,"children":577},{},[578],{"type":45,"value":579},"Azure-hosted",{"type":39,"tag":289,"props":581,"children":582},{},[583],{"type":39,"tag":54,"props":584,"children":586},{"className":585},[],[587],{"type":45,"value":588},"system-type-azure",{"type":39,"tag":289,"props":590,"children":591},{},[592],{"type":45,"value":593},"Azure compute, identity, networking, managed services",{"type":39,"tag":48,"props":595,"children":596},{},[597,602],{"type":39,"tag":66,"props":598,"children":599},{},[600],{"type":45,"value":601},"Design philosophies",{"type":45,"value":603}," -- which does the system claim or embody?",{"type":39,"tag":254,"props":605,"children":606},{},[607,626],{"type":39,"tag":258,"props":608,"children":609},{},[610],{"type":39,"tag":262,"props":611,"children":612},{},[613,618,622],{"type":39,"tag":266,"props":614,"children":615},{},[616],{"type":45,"value":617},"Philosophy",{"type":39,"tag":266,"props":619,"children":620},{},[621],{"type":45,"value":275},{"type":39,"tag":266,"props":623,"children":624},{},[625],{"type":45,"value":280},{"type":39,"tag":282,"props":627,"children":628},{},[629,651,673],{"type":39,"tag":262,"props":630,"children":631},{},[632,637,646],{"type":39,"tag":289,"props":633,"children":634},{},[635],{"type":45,"value":636},"Linux\u002FUnix",{"type":39,"tag":289,"props":638,"children":639},{},[640],{"type":39,"tag":54,"props":641,"children":643},{"className":642},[],[644],{"type":45,"value":645},"design-philosophy-linux",{"type":39,"tag":289,"props":647,"children":648},{},[649],{"type":45,"value":650},"Mechanism vs policy, composability, small sharp tools",{"type":39,"tag":262,"props":652,"children":653},{},[654,659,668],{"type":39,"tag":289,"props":655,"children":656},{},[657],{"type":45,"value":658},"Domain-driven",{"type":39,"tag":289,"props":660,"children":661},{},[662],{"type":39,"tag":54,"props":663,"children":665},{"className":664},[],[666],{"type":45,"value":667},"design-philosophy-domain-driven",{"type":39,"tag":289,"props":669,"children":670},{},[671],{"type":45,"value":672},"Bounded contexts, ubiquitous language, aggregates",{"type":39,"tag":262,"props":674,"children":675},{},[676,681,690],{"type":39,"tag":289,"props":677,"children":678},{},[679],{"type":45,"value":680},"Object-oriented",{"type":39,"tag":289,"props":682,"children":683},{},[684],{"type":39,"tag":54,"props":685,"children":687},{"className":686},[],[688],{"type":45,"value":689},"design-philosophy-object-oriented",{"type":39,"tag":289,"props":691,"children":692},{},[693],{"type":45,"value":694},"SOLID, composition over inheritance, protocols\u002Ftraits",{"type":39,"tag":48,"props":696,"children":697},{},[698],{"type":39,"tag":66,"props":699,"children":700},{},[701],{"type":45,"value":702},"After classifying, immediately load ALL matching skills:",{"type":39,"tag":82,"props":704,"children":707},{"className":705,"code":706,"language":45},[85],"load_skill(skill_name=\"system-type-event-driven\")\nload_skill(skill_name=\"system-type-cli-tool\")\nload_skill(skill_name=\"design-philosophy-linux\")\n# ... every skill that matches\n",[708],{"type":39,"tag":54,"props":709,"children":710},{"__ignoreMap":90},[711],{"type":45,"value":706},{"type":39,"tag":48,"props":713,"children":714},{},[715,717],{"type":45,"value":716},"Present the classification to the user: ",{"type":39,"tag":66,"props":718,"children":719},{},[720,722,728,730,735],{"type":45,"value":721},"\"I've classified this system as ",{"type":39,"tag":723,"props":724,"children":725},"span",{},[726],{"type":45,"value":727},"types",{"type":45,"value":729},". I've loaded domain skills for ",{"type":39,"tag":723,"props":731,"children":732},{},[733],{"type":45,"value":734},"list",{"type":45,"value":736},". Does this classification match your understanding? Anything I should add?\"",{"type":39,"tag":48,"props":738,"children":739},{},[740,745],{"type":39,"tag":66,"props":741,"children":742},{},[743],{"type":45,"value":744},"Why this matters:",{"type":45,"value":746}," Domain skills contain failure modes, anti-patterns, and evaluation criteria specific to each system type. Without them, the review produces generic findings. With them, findings are grounded in domain-specific patterns -- \"this violates event-driven reliability patterns\" rather than \"this might have a timeout issue.\"",{"type":39,"tag":141,"props":748,"children":750},{"id":749},"step-3-evaluate-against-the-codebase",[751],{"type":45,"value":752},"Step 3: Evaluate Against the Codebase",{"type":39,"tag":48,"props":754,"children":755},{},[756],{"type":45,"value":757},"If reviewing an existing system or a design that modifies one:",{"type":39,"tag":186,"props":759,"children":760},{},[761,790,795],{"type":39,"tag":157,"props":762,"children":763},{},[764,766,772,774,780,782,788],{"type":45,"value":765},"Use ",{"type":39,"tag":54,"props":767,"children":769},{"className":768},[],[770],{"type":45,"value":771},"read_file",{"type":45,"value":773},", ",{"type":39,"tag":54,"props":775,"children":777},{"className":776},[],[778],{"type":45,"value":779},"grep",{"type":45,"value":781},", and ",{"type":39,"tag":54,"props":783,"children":785},{"className":784},[],[786],{"type":45,"value":787},"LSP",{"type":45,"value":789}," to examine the actual code",{"type":39,"tag":157,"props":791,"children":792},{},[793],{"type":45,"value":794},"Compare the design's claims against reality",{"type":39,"tag":157,"props":796,"children":797},{},[798],{"type":45,"value":799},"Identify gaps between the design document and the implementation",{"type":39,"tag":48,"props":801,"children":802},{},[803,808],{"type":39,"tag":66,"props":804,"children":805},{},[806],{"type":45,"value":807},"Implementation viability check:",{"type":45,"value":809}," If the design proposes changes to an existing system, trace the impact exhaustively:",{"type":39,"tag":186,"props":811,"children":812},{},[813,834,839],{"type":39,"tag":157,"props":814,"children":815},{},[816,818,824,826,832],{"type":45,"value":817},"Use LSP ",{"type":39,"tag":54,"props":819,"children":821},{"className":820},[],[822],{"type":45,"value":823},"findReferences",{"type":45,"value":825}," and ",{"type":39,"tag":54,"props":827,"children":829},{"className":828},[],[830],{"type":45,"value":831},"incomingCalls",{"type":45,"value":833}," to find every usage of components being changed",{"type":39,"tag":157,"props":835,"children":836},{},[837],{"type":45,"value":838},"For each usage, validate the new design can replace the old seamlessly",{"type":39,"tag":157,"props":840,"children":841},{},[842],{"type":45,"value":843},"Flag any callsite where the replacement is non-trivial or ambiguous",{"type":39,"tag":48,"props":845,"children":846},{},[847],{"type":45,"value":848},"This is not optional exploration -- it is the mechanism that prevents \"it worked in theory but broke 12 callsites\" surprises.",{"type":39,"tag":141,"props":850,"children":852},{"id":851},"step-4-adversarial-analysis",[853],{"type":45,"value":854},"Step 4: Adversarial Analysis",{"type":39,"tag":48,"props":856,"children":857},{},[858],{"type":45,"value":859},"Delegate the deep adversarial review rather than doing it inline:",{"type":39,"tag":48,"props":861,"children":862},{},[863],{"type":39,"tag":66,"props":864,"children":865},{},[866],{"type":45,"value":867},"Option A -- Delegate to the systems-design-critic agent:",{"type":39,"tag":82,"props":869,"children":872},{"className":870,"code":871,"language":45},[85],"delegate(\n  agent=\"systems-design:systems-design-critic\",\n  instruction=\"Review this design: [design content or file path]. System context: [relevant codebase findings from Step 3]. System classification: [types from Step 2]\",\n  context_depth=\"recent\",\n  context_scope=\"agents\"\n)\n",[873],{"type":39,"tag":54,"props":874,"children":875},{"__ignoreMap":90},[876],{"type":45,"value":871},{"type":39,"tag":48,"props":878,"children":879},{},[880],{"type":39,"tag":66,"props":881,"children":882},{},[883],{"type":45,"value":884},"Option B -- Use the adversarial-review skill (spawns 6 parallel agents):",{"type":39,"tag":82,"props":886,"children":889},{"className":887,"code":888,"language":45},[85],"load_skill(skill_name=\"adversarial-review\")\n",[890],{"type":39,"tag":54,"props":891,"children":892},{"__ignoreMap":90},[893],{"type":45,"value":888},{"type":39,"tag":48,"props":895,"children":896},{},[897],{"type":45,"value":898},"Choose Option A when you have accumulated codebase context to pass along. Choose Option B for a quick parallel review when the design is self-contained.",{"type":39,"tag":48,"props":900,"children":901},{},[902,907],{"type":39,"tag":66,"props":903,"children":904},{},[905],{"type":45,"value":906},"Pass the system classification to the agent.",{"type":45,"value":908}," The critic should know what type of system it is reviewing so it can apply domain-specific failure modes rather than generic concerns.",{"type":39,"tag":141,"props":910,"children":912},{"id":911},"step-5-tradeoff-validation",[913],{"type":45,"value":914},"Step 5: Tradeoff Validation",{"type":39,"tag":48,"props":916,"children":917},{},[918],{"type":45,"value":919},"After the adversarial analysis returns, load the tradeoff-analysis skill if the review reveals unclear tradeoffs:",{"type":39,"tag":82,"props":921,"children":924},{"className":922,"code":923,"language":45},[85],"load_skill(skill_name=\"tradeoff-analysis\")\n",[925],{"type":39,"tag":54,"props":926,"children":927},{"__ignoreMap":90},[928],{"type":45,"value":923},{"type":39,"tag":48,"props":930,"children":931},{},[932],{"type":45,"value":933},"Verify:",{"type":39,"tag":186,"props":935,"children":936},{},[937,942,947],{"type":39,"tag":157,"props":938,"children":939},{},[940],{"type":45,"value":941},"Are the stated tradeoffs actually what the design achieves?",{"type":39,"tag":157,"props":943,"children":944},{},[945],{"type":45,"value":946},"Are there unstated tradeoffs the designers may not have recognized?",{"type":39,"tag":157,"props":948,"children":949},{},[950],{"type":45,"value":951},"Does the design optimize for what the team actually needs?",{"type":39,"tag":48,"props":953,"children":954},{},[955],{"type":39,"tag":66,"props":956,"children":957},{},[958],{"type":45,"value":959},"Apply the Simplicity and DX lenses:",{"type":39,"tag":186,"props":961,"children":962},{},[963,968,973,978],{"type":39,"tag":157,"props":964,"children":965},{},[966],{"type":45,"value":967},"Isn't there a simpler way to accomplish what this design does? What concept have we uncovered that we didn't know before?",{"type":39,"tag":157,"props":969,"children":970},{},[971],{"type":45,"value":972},"Are we mixing two concerns here? Aren't these actually separate responsibilities?",{"type":39,"tag":157,"props":974,"children":975},{},[976],{"type":45,"value":977},"How many concepts must a developer hold in their head to use this correctly? Is the naming consistent and clear?",{"type":39,"tag":157,"props":979,"children":980},{},[981],{"type":45,"value":982},"Does this diverge from ecosystem conventions? If so, is the divergence justified?",{"type":39,"tag":141,"props":984,"children":986},{"id":985},"step-6-synthesize-and-recommend",[987],{"type":45,"value":988},"Step 6: Synthesize and Recommend",{"type":39,"tag":48,"props":990,"children":991},{},[992],{"type":45,"value":993},"Present findings to the user.",{"type":39,"tag":48,"props":995,"children":996},{},[997],{"type":39,"tag":66,"props":998,"children":999},{},[1000],{"type":45,"value":1001},"Before presenting, categorize each finding:",{"type":39,"tag":186,"props":1003,"children":1004},{},[1005,1015],{"type":39,"tag":157,"props":1006,"children":1007},{},[1008,1013],{"type":39,"tag":66,"props":1009,"children":1010},{},[1011],{"type":45,"value":1012},"Architectural",{"type":45,"value":1014}," -- boundaries, data flow, failure modes, coupling, scalability, consistency model, state ownership. These belong in a systems-design review.",{"type":39,"tag":157,"props":1016,"children":1017},{},[1018,1023],{"type":39,"tag":66,"props":1019,"children":1020},{},[1021],{"type":45,"value":1022},"Implementation",{"type":45,"value":1024}," -- code quality, specific function issues, API ergonomics, type safety, naming. These are useful but secondary.",{"type":39,"tag":48,"props":1026,"children":1027},{},[1028],{"type":45,"value":1029},"If more than 30% of findings are implementation-level, you have drifted from the systems-design scope. Reframe implementation findings in architectural terms or separate them into a distinct section.",{"type":39,"tag":48,"props":1031,"children":1032},{},[1033],{"type":39,"tag":66,"props":1034,"children":1035},{},[1036],{"type":45,"value":1037},"Check design integrity before finalizing:",{"type":39,"tag":186,"props":1039,"children":1040},{},[1041,1046,1051],{"type":39,"tag":157,"props":1042,"children":1043},{},[1044],{"type":45,"value":1045},"Is this design consistent with previous design decisions in the system? If not, is the inconsistency explicitly reconciled?",{"type":39,"tag":157,"props":1047,"children":1048},{},[1049],{"type":45,"value":1050},"Who owns this component? Does it really belong here, or are we changing the responsibility of something established?",{"type":39,"tag":157,"props":1052,"children":1053},{},[1054],{"type":45,"value":1055},"Does this introduce a second way to do something the system already does?",{"type":39,"tag":48,"props":1057,"children":1058},{},[1059],{"type":39,"tag":66,"props":1060,"children":1061},{},[1062],{"type":45,"value":1063},"Assessment Summary:",{"type":39,"tag":186,"props":1065,"children":1066},{},[1067,1077,1086],{"type":39,"tag":157,"props":1068,"children":1069},{},[1070,1072],{"type":45,"value":1071},"Overall health: ",{"type":39,"tag":723,"props":1073,"children":1074},{},[1075],{"type":45,"value":1076},"strong \u002F has concerns \u002F needs significant revision",{"type":39,"tag":157,"props":1078,"children":1079},{},[1080,1082],{"type":45,"value":1081},"Critical risks: ",{"type":39,"tag":723,"props":1083,"children":1084},{},[1085],{"type":45,"value":734},{"type":39,"tag":157,"props":1087,"children":1088},{},[1089,1091],{"type":45,"value":1090},"Tradeoff alignment: ",{"type":39,"tag":723,"props":1092,"children":1093},{},[1094],{"type":45,"value":1095},"does the design achieve what it claims?",{"type":39,"tag":48,"props":1097,"children":1098},{},[1099,1104],{"type":39,"tag":66,"props":1100,"children":1101},{},[1102],{"type":45,"value":1103},"Recommendation:",{"type":45,"value":1105}," One of:",{"type":39,"tag":186,"props":1107,"children":1108},{},[1109,1119,1134],{"type":39,"tag":157,"props":1110,"children":1111},{},[1112,1117],{"type":39,"tag":66,"props":1113,"children":1114},{},[1115],{"type":45,"value":1116},"Proceed",{"type":45,"value":1118}," -- risks are acceptable, design is sound",{"type":39,"tag":157,"props":1120,"children":1121},{},[1122,1127,1129],{"type":39,"tag":66,"props":1123,"children":1124},{},[1125],{"type":45,"value":1126},"Proceed with modifications",{"type":45,"value":1128}," -- ",{"type":39,"tag":723,"props":1130,"children":1131},{},[1132],{"type":45,"value":1133},"specific changes needed",{"type":39,"tag":157,"props":1135,"children":1136},{},[1137,1142,1143],{"type":39,"tag":66,"props":1138,"children":1139},{},[1140],{"type":45,"value":1141},"Reconsider",{"type":45,"value":1128},{"type":39,"tag":723,"props":1144,"children":1145},{},[1146],{"type":45,"value":1147},"fundamental issue that requires rethinking",{"type":39,"tag":48,"props":1149,"children":1150},{},[1151,1152],{"type":45,"value":218},{"type":39,"tag":66,"props":1153,"children":1154},{},[1155],{"type":45,"value":1156},"\"Do these findings match your intuition? Are there aspects of the design I should look at more closely?\"",{"type":39,"tag":141,"props":1158,"children":1160},{"id":1159},"step-7-capture-action-items",[1161],{"type":45,"value":1162},"Step 7: Capture Action Items",{"type":39,"tag":48,"props":1164,"children":1165},{},[1166],{"type":45,"value":1167},"If the user wants to proceed with modifications, produce a clear action list:",{"type":39,"tag":153,"props":1169,"children":1170},{},[1171,1176,1181],{"type":39,"tag":157,"props":1172,"children":1173},{},[1174],{"type":45,"value":1175},"What to change (specific, not vague)",{"type":39,"tag":157,"props":1177,"children":1178},{},[1179],{"type":45,"value":1180},"Why (linked to a specific risk or finding)",{"type":39,"tag":157,"props":1182,"children":1183},{},[1184],{"type":45,"value":1185},"Priority (must-do vs should-do vs nice-to-do)",{"type":39,"tag":99,"props":1187,"children":1189},{"id":1188},"handling-follow-up-questions-after-agent-work",[1190],{"type":45,"value":1191},"Handling Follow-Up Questions After Agent Work",{"type":39,"tag":48,"props":1193,"children":1194},{},[1195],{"type":45,"value":1196},"When agents have already completed analysis and the user asks a follow-up question, do NOT re-delegate from scratch. Prior agent work is expensive -- treat it as an asset, not a disposable draft.",{"type":39,"tag":48,"props":1198,"children":1199},{},[1200],{"type":39,"tag":66,"props":1201,"children":1202},{},[1203],{"type":45,"value":1204},"Decision rule:",{"type":39,"tag":153,"props":1206,"children":1207},{},[1208,1218,1228],{"type":39,"tag":157,"props":1209,"children":1210},{},[1211,1216],{"type":39,"tag":66,"props":1212,"children":1213},{},[1214],{"type":45,"value":1215},"Can I answer from prior agent findings?",{"type":45,"value":1217}," If the architect or critic already analyzed this system, synthesize their findings into a response. This is conversation work, not analysis work -- it belongs to you.",{"type":39,"tag":157,"props":1219,"children":1220},{},[1221,1226],{"type":39,"tag":66,"props":1222,"children":1223},{},[1224],{"type":45,"value":1225},"Does the follow-up require genuinely new analysis?",{"type":45,"value":1227}," New scope, new data, or a different analytical frame that prior work didn't cover. Only then re-delegate.",{"type":39,"tag":157,"props":1229,"children":1230},{},[1231,1236,1238,1244],{"type":39,"tag":66,"props":1232,"children":1233},{},[1234],{"type":45,"value":1235},"If re-delegating, build on prior work.",{"type":45,"value":1237}," Resume the existing agent session (",{"type":39,"tag":54,"props":1239,"children":1241},{"className":1240},[],[1242],{"type":45,"value":1243},"session_id",{"type":45,"value":1245},") when possible. If spawning fresh, include a summary of prior findings in the instruction so the agent extends rather than repeats.",{"type":39,"tag":48,"props":1247,"children":1248},{},[1249,1254],{"type":39,"tag":66,"props":1250,"children":1251},{},[1252],{"type":45,"value":1253},"The test:",{"type":45,"value":1255}," If the new delegation instruction overlaps >50% with a prior one, you are re-doing work. Synthesize instead.",{"type":39,"tag":48,"props":1257,"children":1258},{},[1259,1264],{"type":39,"tag":66,"props":1260,"children":1261},{},[1262],{"type":45,"value":1263},"Anti-pattern:",{"type":45,"value":1265}," User asks about higher-level architectural decisions. You feel uncertain about your own synthesis, so you re-spawn the architect and critic from scratch. Both agents produce findings that overlap heavily with their first pass. Hours of computation wasted, and the user watches agents return similar results twice.",{"type":39,"tag":48,"props":1267,"children":1268},{},[1269,1274],{"type":39,"tag":66,"props":1270,"children":1271},{},[1272],{"type":45,"value":1273},"Correct pattern:",{"type":45,"value":1275}," User asks about architectural decisions. You check: the architect already assessed boundaries, coupling, and failure modes; the critic already reviewed from 6 perspectives. You synthesize those findings into an answer focused on the user's specific question. If the question touches something genuinely uncovered (e.g., a specific subsystem the prior analysis didn't examine), you delegate only that gap -- and reference prior findings so the agent builds on them.",{"type":39,"tag":48,"props":1277,"children":1278},{},[1279,1284],{"type":39,"tag":66,"props":1280,"children":1281},{},[1282],{"type":45,"value":1283},"The \"analysis vs conversation\" heuristic:",{"type":45,"value":1285}," If answering requires reasoning about tradeoffs, failure modes, or design alternatives using information you don't have, it is ANALYSIS -- delegate it. If answering requires presenting, reframing, or connecting findings that agents already produced, it is CONVERSATION -- do it yourself. When in doubt, check whether prior agent results already contain the answer before delegating.",{"items":1287,"total":1361},[1288,1299,1311,1320,1331,1345,1355],{"slug":667,"name":667,"fn":1289,"description":1290,"org":1291,"tags":1292,"stars":22,"repoUrl":23,"updatedAt":1298},"model complex systems with domain-driven design","Domain-Driven Design as a lens for system architecture — bounded contexts, aggregates, ubiquitous language, context mapping, domain events, and strategic vs tactical patterns. Use when modeling complex business domains, defining service boundaries, or evaluating whether a system's structure reflects its domain.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1293,1294,1297],{"name":13,"slug":14,"type":15},{"name":1295,"slug":1296,"type":15},"Data Modeling","data-modeling",{"name":17,"slug":18,"type":15},"2026-07-07T06:53:28.678913",{"slug":645,"name":645,"fn":1300,"description":1301,"org":1302,"tags":1303,"stars":22,"repoUrl":23,"updatedAt":1310},"apply Unix and Linux design philosophy","The Unix\u002FLinux design philosophy as a lens for system design — mechanism vs policy, composability, small tools, text streams, convention over configuration, and the principle of least surprise. Use when evaluating designs for composability, simplicity, or separation of concerns.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1304,1305,1306,1309],{"name":13,"slug":14,"type":15},{"name":20,"slug":21,"type":15},{"name":1307,"slug":1308,"type":15},"Linux","linux",{"name":17,"slug":18,"type":15},"2026-05-13T06:14:50.770309",{"slug":478,"name":478,"fn":1312,"description":1313,"org":1314,"tags":1315,"stars":22,"repoUrl":23,"updatedAt":1319},"apply distributed systems design patterns","Foundational patterns for distributed systems — consensus, consistency models, replication, partitioning, clock synchronization, distributed transactions, and failure modes. Use when designing or evaluating any system that spans multiple nodes, processes, or failure domains.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1316,1317,1318],{"name":13,"slug":14,"type":15},{"name":20,"slug":21,"type":15},{"name":17,"slug":18,"type":15},"2026-05-13T06:14:49.388552",{"slug":324,"name":324,"fn":1321,"description":1322,"org":1323,"tags":1324,"stars":22,"repoUrl":23,"updatedAt":1330},"design event-driven and message-based systems","Domain patterns for event-driven and message-based systems — pub\u002Fsub, event sourcing, CQRS, sagas, delivery guarantees, schema evolution, and failure modes. Use when designing or evaluating systems built around events, messages, or asynchronous workflows.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1325,1326,1329],{"name":13,"slug":14,"type":15},{"name":1327,"slug":1328,"type":15},"Messaging","messaging",{"name":17,"slug":18,"type":15},"2026-07-07T06:53:27.291427",{"slug":456,"name":456,"fn":1332,"description":1333,"org":1334,"tags":1335,"stars":22,"repoUrl":23,"updatedAt":1344},"design ML serving and training systems","Domain patterns for ML\u002FAI serving and training systems — model serving, feature stores, training pipelines, experiment tracking, A\u002FB testing, GPU scheduling, and failure modes. Use when designing or evaluating machine learning infrastructure, model serving platforms, or AI-powered product features.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1336,1339,1342,1343],{"name":1337,"slug":1338,"type":15},"AI Infrastructure","ai-infrastructure",{"name":1340,"slug":1341,"type":15},"Deep Learning","deep-learning",{"name":20,"slug":21,"type":15},{"name":17,"slug":18,"type":15},"2026-07-03T16:31:59.997224",{"slug":412,"name":412,"fn":1346,"description":1347,"org":1348,"tags":1349,"stars":22,"repoUrl":23,"updatedAt":1354},"design real-time and collaborative systems","Domain patterns for real-time and collaborative systems — persistent connections, state synchronization, conflict resolution, presence, fan-out, and failure modes. Use when designing or evaluating chat systems, collaborative editors, live dashboards, gaming backends, or any system with bidirectional real-time communication.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1350,1351,1353],{"name":13,"slug":14,"type":15},{"name":403,"slug":1352,"type":15},"real-time",{"name":17,"slug":18,"type":15},"2026-07-07T06:53:26.004577",{"slug":4,"name":4,"fn":5,"description":6,"org":1356,"tags":1357,"stars":22,"repoUrl":23,"updatedAt":24},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1358,1359,1360],{"name":13,"slug":14,"type":15},{"name":20,"slug":21,"type":15},{"name":17,"slug":18,"type":15},7,{"items":1363,"total":1554},[1364,1384,1405,1426,1439,1456,1467,1480,1495,1510,1529,1542],{"slug":1365,"name":1365,"fn":1366,"description":1367,"org":1368,"tags":1369,"stars":1381,"repoUrl":1382,"updatedAt":1383},"rushstack-best-practices","manage Rush monorepos with best practices","Provides best practices and guidance for working with Rush monorepos. Use when the user is working in a Rush-based repository, asks about Rush commands (install, update, build, rebuild), needs help with project selection, dependency management, build caching, subspace configuration, or troubleshooting Rush-specific issues.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1370,1371,1374,1375,1378],{"name":20,"slug":21,"type":15},{"name":1372,"slug":1373,"type":15},"Local Development","local-development",{"name":9,"slug":8,"type":15},{"name":1376,"slug":1377,"type":15},"Project Management","project-management",{"name":1379,"slug":1380,"type":15},"Rush","rush",6484,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Frushstack","2026-04-06T18:34:44.965032",{"slug":1385,"name":1385,"fn":1386,"description":1387,"org":1388,"tags":1389,"stars":1402,"repoUrl":1403,"updatedAt":1404},"azure-ai-agents-persistent-dotnet","build AI agents with Azure .NET SDK","Azure AI Agents Persistent SDK for .NET. Low-level SDK for creating and managing AI agents with threads, messages, runs, and tools. Use for agent CRUD, conversation threads, streaming responses, function calling, file search, and code interpreter. Triggers: \"PersistentAgentsClient\", \"persistent agents\", \"agent threads\", \"agent runs\", \"streaming agents\", \"function calling agents .NET\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1390,1393,1396,1399],{"name":1391,"slug":1392,"type":15},".NET","net",{"name":1394,"slug":1395,"type":15},"Agents","agents",{"name":1397,"slug":1398,"type":15},"Azure","azure",{"name":1400,"slug":1401,"type":15},"LLM","llm",2804,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fskills","2026-07-03T16:32:10.297433",{"slug":1406,"name":1406,"fn":1407,"description":1408,"org":1409,"tags":1410,"stars":1402,"repoUrl":1403,"updatedAt":1425},"azure-ai-anomalydetector-java","build anomaly detection applications with Java","Build anomaly detection applications with Azure AI Anomaly Detector SDK for Java. Use when implementing univariate\u002Fmultivariate anomaly detection, time-series analysis, or AI-powered monitoring.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1411,1414,1415,1418,1421,1422],{"name":1412,"slug":1413,"type":15},"Analytics","analytics",{"name":1397,"slug":1398,"type":15},{"name":1416,"slug":1417,"type":15},"Data Analysis","data-analysis",{"name":1419,"slug":1420,"type":15},"Java","java",{"name":9,"slug":8,"type":15},{"name":1423,"slug":1424,"type":15},"Monitoring","monitoring","2026-05-13T06:14:16.261754",{"slug":1427,"name":1427,"fn":1428,"description":1429,"org":1430,"tags":1431,"stars":1402,"repoUrl":1403,"updatedAt":1438},"azure-ai-contentsafety-java","build content moderation applications with Azure AI","Build content moderation applications with Azure AI Content Safety SDK for Java. Use when implementing text\u002Fimage analysis, blocklist management, or harm detection for hate, violence, sexual content, and self-harm.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1432,1433,1434,1435],{"name":1337,"slug":1338,"type":15},{"name":1397,"slug":1398,"type":15},{"name":1419,"slug":1420,"type":15},{"name":1436,"slug":1437,"type":15},"Security","security","2026-07-07T06:53:31.293235",{"slug":1440,"name":1440,"fn":1441,"description":1442,"org":1443,"tags":1444,"stars":1402,"repoUrl":1403,"updatedAt":1455},"azure-ai-contentsafety-py","detect harmful content with Azure AI Content Safety","Azure AI Content Safety SDK for Python. Use for detecting harmful content in text and images with multi-severity classification.\nTriggers: \"azure-ai-contentsafety\", \"ContentSafetyClient\", \"content moderation\", \"harmful content\", \"text analysis\", \"image analysis\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1445,1446,1449,1450,1451,1454],{"name":1397,"slug":1398,"type":15},{"name":1447,"slug":1448,"type":15},"Compliance","compliance",{"name":1400,"slug":1401,"type":15},{"name":9,"slug":8,"type":15},{"name":1452,"slug":1453,"type":15},"Python","python",{"name":1436,"slug":1437,"type":15},"2026-07-18T05:14:23.017504",{"slug":1457,"name":1457,"fn":1458,"description":1459,"org":1460,"tags":1461,"stars":1402,"repoUrl":1403,"updatedAt":1466},"azure-ai-language-conversations-py","implement conversational language understanding with Python","Implement Conversational Language Understanding (CLU) using the azure-ai-language-conversations Python SDK. Use when working with ConversationAnalysisClient to analyze conversation intent and entities, building NLP features, or integrating language understanding into applications.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1462,1463,1464,1465],{"name":1412,"slug":1413,"type":15},{"name":1397,"slug":1398,"type":15},{"name":1400,"slug":1401,"type":15},{"name":1452,"slug":1453,"type":15},"2026-07-31T05:54:29.068751",{"slug":1468,"name":1468,"fn":1469,"description":1470,"org":1471,"tags":1472,"stars":1402,"repoUrl":1403,"updatedAt":1479},"azure-ai-translation-text-py","translate text using Azure AI services","Azure AI Text Translation SDK for real-time text translation, transliteration, language detection, and dictionary lookup. Use for translating text content in applications.\nTriggers: \"text translation\", \"translator\", \"translate text\", \"transliterate\", \"TextTranslationClient\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1473,1476,1477,1478],{"name":1474,"slug":1475,"type":15},"API Development","api-development",{"name":1397,"slug":1398,"type":15},{"name":9,"slug":8,"type":15},{"name":1452,"slug":1453,"type":15},"2026-07-18T05:14:16.988376",{"slug":1481,"name":1481,"fn":1482,"description":1483,"org":1484,"tags":1485,"stars":1402,"repoUrl":1403,"updatedAt":1494},"azure-ai-vision-imageanalysis-py","analyze images with Azure AI Vision","Azure AI Vision Image Analysis SDK for captions, tags, objects, OCR, people detection, and smart cropping. Use for computer vision and image understanding tasks.\nTriggers: \"image analysis\", \"computer vision\", \"OCR\", \"object detection\", \"ImageAnalysisClient\", \"image caption\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1486,1487,1490,1493],{"name":1397,"slug":1398,"type":15},{"name":1488,"slug":1489,"type":15},"Computer Vision","computer-vision",{"name":1491,"slug":1492,"type":15},"Images","images",{"name":1452,"slug":1453,"type":15},"2026-07-18T05:14:18.007737",{"slug":1496,"name":1496,"fn":1497,"description":1498,"org":1499,"tags":1500,"stars":1402,"repoUrl":1403,"updatedAt":1509},"azure-appconfiguration-java","manage configuration with Azure App Configuration","Azure App Configuration SDK for Java. Centralized application configuration management with key-value settings, feature flags, and snapshots.\nTriggers: \"ConfigurationClient java\", \"app configuration java\", \"feature flag java\", \"configuration setting java\", \"azure config java\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1501,1502,1505,1508],{"name":1397,"slug":1398,"type":15},{"name":1503,"slug":1504,"type":15},"Configuration","configuration",{"name":1506,"slug":1507,"type":15},"Feature Flags","feature-flags",{"name":1419,"slug":1420,"type":15},"2026-07-03T16:32:01.278468",{"slug":1511,"name":1511,"fn":1512,"description":1513,"org":1514,"tags":1515,"stars":1402,"repoUrl":1403,"updatedAt":1528},"azure-cosmos-rust","build applications with Azure Cosmos DB","Azure Cosmos DB library for Rust (NoSQL API). Document CRUD, containers, and globally distributed data.\nTriggers: \"cosmos db rust\", \"CosmosClient rust\", \"document crud rust\", \"NoSQL rust\", \"partition key rust\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1516,1519,1522,1525],{"name":1517,"slug":1518,"type":15},"Cosmos DB","cosmos-db",{"name":1520,"slug":1521,"type":15},"Database","database",{"name":1523,"slug":1524,"type":15},"NoSQL","nosql",{"name":1526,"slug":1527,"type":15},"Rust","rust","2026-07-31T05:54:27.021432",{"slug":1530,"name":1530,"fn":1512,"description":1531,"org":1532,"tags":1533,"stars":1402,"repoUrl":1403,"updatedAt":1541},"azure-cosmos-ts","Azure Cosmos DB JavaScript\u002FTypeScript SDK (@azure\u002Fcosmos) for data plane operations. Use for CRUD operations on documents, queries, bulk operations, and container management. Triggers: \"Cosmos DB\", \"@azure\u002Fcosmos\", \"CosmosClient\", \"document CRUD\", \"NoSQL queries\", \"bulk operations\", \"partition key\", \"container.items\".\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1534,1535,1536,1537,1538],{"name":1517,"slug":1518,"type":15},{"name":1520,"slug":1521,"type":15},{"name":9,"slug":8,"type":15},{"name":1523,"slug":1524,"type":15},{"name":1539,"slug":1540,"type":15},"TypeScript","typescript","2026-07-03T16:31:19.368382",{"slug":1543,"name":1543,"fn":1544,"description":1545,"org":1546,"tags":1547,"stars":1402,"repoUrl":1403,"updatedAt":1553},"azure-data-tables-java","build table storage applications with Java","Build table storage applications with Azure Tables SDK for Java. Use when working with Azure Table Storage or Cosmos DB Table API for NoSQL key-value data, schemaless storage, or structured data at scale.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1548,1549,1550,1551,1552],{"name":1397,"slug":1398,"type":15},{"name":1517,"slug":1518,"type":15},{"name":1520,"slug":1521,"type":15},{"name":1419,"slug":1420,"type":15},{"name":1523,"slug":1524,"type":15},"2026-05-13T06:14:17.582229",267]