How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
MicrosoftCLAUDE.mdmicrosoft/powercat/CLAUDE.md
Always read DESIGN.md before making any visual or UI decisions. All font choices, colors, spacing, and aesthetic direction are defined there. Do not deviate without explicit user approval. In QA mode, flag any code that does not match DESIGN.md.
MicrosoftCLAUDE.mdmicrosoft/ShadowFrog/claude.md
ShadowFrog is a suite of AI coding agent skills that build and maintain shadow knowledge bases for any codebase. It consists of 7 skills (shadow-frog, shadow-frog-init, shadow-frog-update, shadow-frog-dream, shadow-frog-nap, shadow-frog-meditate, shadow-frog-viewer) and associated hooks. Nap's proposals remain separate from .shadow/ discoveries. This is a distributable skills package — users install it into their own projects via install.sh or install.ps1.
MicrosoftCLAUDE.mdmicrosoft/Employee-Self-Service-Agent-Developer-Kit/dev-specs/ess-nextgen-migration-toolkit/CLAUDE.md
This folder is the specification source of truth for the ESS NextGen Migration Toolkit (Specification-Driven Development). Read AGENTS.md in this folder for the full AI Agent Operating Manual — specification classification, the dependency-based loading model, specification hierarchy, repository invariants, and the implementation algorithm. It is the canonical constitution; this file adds no independent rules. The buildable toolkit (code) lives separately under ../../tools/ess-nextgen-migration-toolkit/.
MicrosoftCLAUDE.mdmicrosoft/Employee-Self-Service-Agent-Developer-Kit/tools/ess-nextgen-migration-toolkit/CLAUDE.md
Read AGENTS.md for full project context — architecture, project structure, implementation guide, conventions, migration rules, and coding standards.
MicrosoftCLAUDE.mdmicrosoft/microsoft-copilot-agents-playbook/04-evaluate/demo-chef-agent/Chef-Agent-Initial/CLAUDE.md
This is an M365 Copilot declarative agent project managed by the ATK CLI. When working on this project, you MUST use the appropriate skill for the task. Do NOT work directly on files without invoking a skill first. This is a declarative agent project. For ANY task in this workspace — regardless of what it is — you MUST invoke the declarative-agent-developer skill BEFORE doing any work. This is not optional. Do not attempt to handle any task yourself. Always…
MicrosoftCLAUDE.mdmicrosoft/Agent365-nodejs/CLAUDE.md
The Microsoft Agent 365 SDK extends the Microsoft 365 Agents SDK with enterprise-grade capabilities for building sophisticated AI agents. This Node.js/TypeScript monorepo provides comprehensive tooling across four core areas: This is a pnpm workspace monorepo with 9 packages in packages/: Dependency Flow: Each package follows this convention: Build System: - Each package builds to both CJS and ESM formats - npm run build runs build:cjs && build:esm - TypeScript uses separate tsconfig files for each module format
MicrosoftCLAUDE.mdmicrosoft/FSQ/CLAUDE.md
This repository uses Spec-Driven Development for project development. Root SPEC.md and relevant module SPEC.md files are the grounding truth for project code and project logic. They do not define the SDD workflow. This file and .github/skills/ control agent write authorization and SDD phase transitions. Project SPEC.md files must not duplicate or override those workflow rules. Ordinary discussion, explanation, review, and planning are read-only. Do not automatically load or invoke repository workflow skills. Before creating, modifying, renaming, or deleting project development…
MicrosoftCLAUDE.mdmicrosoft/mcscatblog/CLAUDE.md
This is the Microsoft Copilot Studio CAT team's technical blog, hosted at https://microsoft.github.io/mcscatblog/. It uses Jekyll with the Chirpy theme for static site generation on GitHub Pages. Posts go in _posts/ with filename format: YYYY-MM-DD-post-slug.md Add new authors to _data/authors.yml with: - name: Display name - email: Optional - avatar: GitHub avatar URL (e.g., https://github.com/username.png) - url: Optional link to profile/website Store post images in assets/posts/<post-slug>/. Reference in posts as /assets/posts/<post-slug>/image.png. See Chirpy documentation for advanced formatting: https://chirpy.cotes.page/posts/write-a-new-post/ Review instructions…
MicrosoftCLAUDE.mdmicrosoft/PowerPlatform-EnterprisePolicies/CLAUDE.md
PowerShell module and scripts for managing Power Platform Enterprise Policies as Azure resources. Supports: - Customer Managed Key (CMK) policies for encryption - Subnet Injection policies for virtual network delegation - Diagnostic tools for VNET functionality troubleshooting Supports multiple Azure environments: AzureCloud, AzureUSGovernment (DoD, USGovHigh), AzureChinaCloud. Testing requirement: Tests must pass on both PowerShell (Core) and Windows PowerShell. The test command only runs in the current environment, so you must invoke it separately in each: CRITICAL: Always verify that Build…
MicrosoftCLAUDE.mdmicrosoft/ppteval/claude-workspace/CLAUDE.md
You are running inside a single-task workspace for the PowerPoint Computer Use benchmark. Each invocation handles exactly one task. The harness has prepared this directory with everything you need. 1. **Read TASK.md and OUTPUT_INSTRUCTIONS.md first.* They define the goal and the exact path you must write your output to. Follow the output path literally — the grader looks for that path and nothing else. 2. *Treat inputs/ as read-only.* Copy the file before modifying it. 3. *Write your result to…
MicrosoftCLAUDE.mdmicrosoft/agent-resources/CLAUDE.md
When the user's request matches an available skill, ALWAYS invoke it using the Skill tool as your FIRST action. Do NOT answer directly, do NOT use other tools first. The skill has specialized workflows that produce better results than ad-hoc answers. Key routing rules: - Product ideas, "is this worth building", brainstorming → invoke office-hours - Bugs, errors, "why is this broken", 500 errors → invoke investigate - Ship, deploy, push, create PR → invoke ship - QA, test the…
MicrosoftCLAUDE.mdmicrosoft/cat/CLAUDE.md
When the user's request matches an available skill, invoke it via the Skill tool. When in doubt, invoke the skill. Key routing rules: - Product ideas/brainstorming → invoke /office-hours - Strategy/scope → invoke /plan-ceo-review - Architecture → invoke /plan-eng-review - Design system/plan review → invoke /design-consultation or /plan-design-review - Full review pipeline → invoke /autoplan - Bugs/errors → invoke /investigate - QA/testing site behavior → invoke /qa or /qa-only - Code review/diff check → invoke /review - Visual polish →…
MicrosoftCLAUDE.mdmicrosoft/AgentPlatformAdvisor/CLAUDE.md
When the user's request matches an available skill, invoke it via the Skill tool. When in doubt, invoke the skill. Key routing rules: - Product ideas/brainstorming -> invoke /office-hours - Strategy/scope -> invoke /plan-ceo-review - Architecture -> invoke /plan-eng-review - Design system/plan review -> invoke /design-consultation or /plan-design-review - Full review pipeline -> invoke /autoplan - Bugs/errors -> invoke /investigate - QA/testing site behavior -> invoke /qa or /qa-only - Code review/diff check -> invoke /review - Visual polish ->…
MicrosoftCLAUDE.mdmicrosoft/ai-cheatcode/CLAUDE.md
"The Cheat Code" is an internal Microsoft newsletter (ABS Tech Strategy) publishing weekly reusable agentic patterns for Copilot Chat. Each issue exists in parallel formats: HTML email, PDF, Viva Amplify content, architecture diagram, and interactive portal page. The PRODUCTION_PLAYBOOK.md is the authoritative reference for all production processes — read it first.
MicrosoftCLAUDE.mdmicrosoft/omnichannel-amsclient/CLAUDE.md
Two-mode client selected by framedMode config flag: FramedClient loads a hidden iframe ({baseUrl}/{version}/iframe.html) containing IframeCommunicator. All AMS API calls execute inside the iframe. Host ↔ iframe communication via postMessage with request/response pattern and UUID-based request correlation. FramedlessClient calls API.ts functions directly via fetch. API layer (API.ts): Five functions — skypeTokenAuth, createObject, uploadDocument, getViewStatus, getView. Auth via skype_token header. AMS endpoint from chatToken.amsEndpoint || chatToken.regionGTMS?.ams.
MicrosoftCLAUDE.mdmicrosoft/amplifier-app-nanoclaw/container/CLAUDE.md
You are a NanoClaw agent. Your name, destinations, and message-sending rules are provided in the runtime system prompt at the top of each turn. Be concise — every message costs the reader's attention. Prefer outcomes over play-by-play; when the work is done, the final message should be about the result, not a transcript of what you did. Files you create are saved in /workspace/agent/. Use this for notes, research, or anything that should persist across turns in this group. The…
MicrosoftCLAUDE.mdmicrosoft/Dev-Trio/CLAUDE.md
This workspace uses the Dev-Trio three-agent methodology. - dt-planner — orchestrates all tasks, delegates to implementer and critic subagents - dt-implementer — applies changes, runs gates - dt-critic — audits changes, issues verdict
MicrosoftCLAUDE.mdmicrosoft/read_for_you/CLAUDE.md
Read for You is a web application that extracts text and multimedia content from PDF documents using Azure Document Intelligence (Intelligent OCR), displays them in an accessible format, and provides text-to-speech (TTS) capabilities for reading aloud. Configure these in frontend/readforyou/src/constants.js: Backend service URLs in backend/readforyou/settings.py: Critical: The TTS implementation uses a sophisticated strategy documented in frontend/readforyou/docs/TTS_Strategy.md. Key features: State management (ReadingPage.vue): - audioMap: Map storing elementIndex → audioBlob - isLoadingAudio: Loading state for UI feedback - hasUsedTTS: Tracks if user…
MicrosoftCLAUDE.mdmicrosoft/bobpgsqldemos/caldovalogistics/demo4-foundry-agent/caldova-logistics-agent/CLAUDE.md
This project uses AGENTS.md as the single source of truth for coding agent instructions. The import below loads it into Claude Code's context.
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.
