MicrosoftCLAUDE.mdvscode-mssql
microsoft/vscode-mssql/CLAUDE.md
See AGENTS.md.
CLAUDE.md examples from real projects
How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
Most stars first · from page 2Worked for most · soon
MicrosoftCLAUDE.mdMicrosoftCLAUDE.mdrnx-kit / fork-sync
microsoft/rnx-kit/incubator/fork-sync/CLAUDE.md
MANDATORY RULES 1. DISCUSS before implementing - Answer questions first, code after agreement 2. Use node directly for TypeScript - NOT tsx, ts-node, or tsc 3. Check the platform - Claude Code knows the OS (Platform: in system prompt). Follow the platform-specific rules below. Sync forked/vendored dependencies with upstream using Git 3-way merge with AI-assisted conflict resolution. Note: Claude/Copilot integration tests are slow (invoke external CLIs) and are skipped by default. Use yarn test:all or set RUN_INTEGRATION=1 to include them.
MicrosoftCLAUDE.mdplaywright-java
microsoft/playwright-java/CLAUDE.md
The Java client is a port of the JavaScript client in ../playwright/packages/playwright-core/src/client/. When implementing or changing a method, read the corresponding JS file first and mirror its logic. Project checkouts (including the upstream playwright repo) live in the parent directory (../). Use the gh cli to interact with GitHub. Semantic commit messages: label(scope): description Labels: fix, feat, chore, docs, test, devops Never add Co-Authored-By agents in commit message. Never add "Generated with" in commit message. Never add test plan to…
MicrosoftCLAUDE.mdfluentui-react-native
microsoft/fluentui-react-native/CLAUDE.md
See @AGENTS.md for guidance on working with code in this repository.
MicrosoftCLAUDE.mdfhir-server
microsoft/fhir-server/CLAUDE.md
Core project guidelines, coding conventions, and architectural principles are in AGENTS.md — read it first. This file adds Claude-specific skill activation instructions. Skills live in .github/skills/<skill-name>/SKILL.md. Claude cannot discover them automatically — use the Read tool to load the relevant skill before starting any architecture task. Example: Read(".github/skills/create-adr/SKILL.md") Happy coding!
MicrosoftCLAUDE.mdskills-for-fabric
microsoft/skills-for-fabric/CLAUDE.md
Updates: fabric-collection is a third-party marketplace, so enable auto-update once: run /plugin, open Marketplaces, select fabric-collection, and choose Enable auto-update. Administrators can instead set "autoUpdate": true on its extraKnownMarketplaces entry in managed settings. To update on demand, run claude plugin update <plugin>@fabric-collection. If these files were copied in loosely rather than installed as a plugin, compare the local package.json version against the remote (git fetch origin main --quiet && git show origin/main:package.json) and re-copy if it is newer. This project…
MicrosoftCLAUDE.mdAIOpsLab
microsoft/AIOpsLab/CLAUDE.md
AIOpsLab is a holistic framework for evaluating autonomous AIOps agents in interactive cloud environments. It orchestrates microservice deployments, fault injection, workload generation, and telemetry collection to benchmark AI agents on detection, localization, analysis, and mitigation tasks. Central Coordination Flow: Key Components: - Orchestrator (aiopslab/orchestrator/orchestrator.py): Central coordinator managing lifecycle from initialization to evaluation - Session (aiopslab/session.py): Tracks agent-environment interaction history, timing, results; supports JSON export and W&B logging - ResponseParser (aiopslab/orchestrator/parser.py): Extracts API calls from agent responses using AST parsing for…
MicrosoftCLAUDE.mdpower-platform-skills / power-automate
microsoft/power-platform-skills/plugins/power-automate/CLAUDE.md
This is the marketplace plugin wrapper for FlowAgent (Power Automate cloud flow tooling). Flow-definition rules, the two-provider auth model (Azure CLI + MSAL), the error reference, and known issues are documented in references/ — see definition-reference.md, connection-patterns.md, and error-troubleshooting.md. Read those before building or editing flows. This file only covers plugin-local specifics; see AGENTS.md (this folder) for the tool-routing rule. - skills/ — verb-first user-invocable skills (create-flow, build-flow, debug-flow, diagnose-flow, manage-flows, browse-flows, manage-desktop-flows, route-environments, setup) - references/ — shared docs…
MicrosoftCLAUDE.mdfabric-toolbox / MicrosoftFabricMgmt
microsoft/fabric-toolbox/tools/MicrosoftFabricMgmt/CLAUDE.md
This document provides comprehensive instructions for AI-assisted development of the MicrosoftFabricMgmt PowerShell module located in tools/MicrosoftFabricMgmt/. This module provides a robust interface for managing Microsoft Fabric resources via the Fabric API. Module Purpose: Enterprise-grade PowerShell module for Microsoft Fabric resource management with excellent error handling, comprehensive logging, and outstanding documentation. Build System: Sampler - Modern PowerShell module scaffolding and build framework Key Frameworks: - PSFramework: Configuration management and logging (Documentation) - Pester: Unit and integration testing - PlatyPS: Help documentation…
MicrosoftCLAUDE.mdfabric-toolbox / MicrosoftFabricMgmtMCPServer
microsoft/fabric-toolbox/tools/MicrosoftFabricMgmtMCPServer/CLAUDE.md
This directory contains a Python-based Model Context Protocol (MCP) server that wraps the MicrosoftFabricMgmt PowerShell module and exposes its capabilities to AI assistants (Claude, GitHub Copilot, etc.). DO NOT edit the PowerShell module source here. All PowerShell module development belongs in ../MicrosoftFabricMgmt/source/ and must follow the module's own CLAUDE.md. Key principle: Python is a thin protocol-translation layer only. All Fabric API logic lives in the PowerShell module. Python does not call Fabric REST APIs directly. Maintains a single long-lived pwsh…
MicrosoftCLAUDE.mdCopilotStudioSamples
microsoft/CopilotStudioSamples/CLAUDE.md
Microsoft Copilot Studio samples repo with a Just the Docs (Jekyll) site for navigation and search. Live site: https://microsoft.github.io/CopilotStudioSamples/ Solutions follow the PnP format: Front matter: README should include: - What the solution does and which Copilot Studio features it demonstrates - Screenshots in assets/ (reference as ) - Installation steps (import zip via make.powerapps.com > Solutions > Import) - Any connection references or environment variables to configure - Known issues or limitations Update authoring/solutions/README.md Contents table after adding. Add…
MicrosoftCLAUDE.mdTypeAgent / ts
microsoft/TypeAgent/ts/CLAUDE.md
A fresh checkout of this repo — a new clone, a git worktree (each Copilot CLI session gets its own), or a cloud/container workspace — has the committed files but not the gitignored files that let the repo install and run. Those must be provisioned once. From ts/: On a brand-new machine where pnpm isn't cached yet, pnpm run setup can't bootstrap itself (corepack would try to fetch pnpm from the blocked public registry). Start with the pnpm-free entrypoint instead…
MicrosoftCLAUDE.mdTypeAgent / planValidationDemo
microsoft/TypeAgent/ts/examples/planValidationDemo/CLAUDE.md
This project is governed by a plan-validation MCP server. *Every file access and shell call you make must go through the validated_ tools.** The built-in Read, Write, Edit, Bash, Glob, Grep tools are NOT available in this session — only the MCP tools provided by the plan-validation server. A live dashboard is running at http://localhost:3100 — every tool call you make is visible there, including blocks. Call getplanschema. Read it carefully — the AgentPlan DSL has strict shape requirements and…
MicrosoftCLAUDE.mdTypeAgent / powershell
microsoft/TypeAgent/ts/packages/agents/powershell/CLAUDE.md
PowerShell captures and reuses PowerShell scripts from reasoning traces. Scripts persist in instance storage (~/.typeagent/profiles/<profile>/powershell/) across sessions. Grammar rules are registered dynamically at runtime — no build step needed for user-created flows. scripts/compileRecipes.mjs reads pending/*.recipe.json and generates build-time artifacts. This is useful for testing but not used in production — production flows are loaded from instance storage at runtime.
MicrosoftCLAUDE.mdTypeAgent / taskflow
microsoft/TypeAgent/ts/packages/agents/taskflow/CLAUDE.md
TaskFlow runs user-taught multi-step macros over existing TypeAgent actions. Flows are pure JSON — no TypeScript logic. A deterministic compile script handles all boilerplate generation (TypeScript type, grammar rule, manifest entry). When asked to "compile pending recipes" or similar: That's it. scripts/compileRecipes.mjs reads pending/*.recipe.json and: If the build fails, fix the error in the generated files and re-run pnpm run build. When asked to write a recipe for a new flow (or if you notice a missing flow): Do NOT…
MicrosoftCLAUDE.mdsemantic-link-labs
microsoft/semantic-link-labs/CLAUDE.md
No summary in the file. Open it to read it.
MicrosoftCLAUDE.mdWindowsProtocolTestSuites
microsoft/WindowsProtocolTestSuites/CLAUDE.md
This is the Windows Protocol Test Suites repository — a .NET-based interoperability testing framework for Microsoft Open Specification protocols. Follow these steps before starting any task: Use kebab-case for the…
MicrosoftCLAUDE.mdplaywright-vscode
microsoft/playwright-vscode/CLAUDE.md
Semantic commit messages: label(scope): description Labels: fix, feat, chore, docs, test, devops Never add Co-Authored-By agents in commit message. Never add "Generated with" in commit message. Never add test plan to PR description. Keep PR description short — a few bullet points at most. Branch naming for issue fixes: fix-<issue-number> Never git push without an explicit instruction to push. Applies even when a PR is already open for the branch — additional commits are immediately visible to reviewers. Commit locally,…
MicrosoftCLAUDE.mdconductor
microsoft/conductor/CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with this repository.
MicrosoftCLAUDE.mdconductor / providers
microsoft/conductor/docs/providers/claude.md
The Claude provider enables Conductor workflows to use Anthropic's Claude models via Pydantic AI (pydantic-ai package, AnthropicModel). The Claude provider delegates its agentic loop, tool execution, and structured output processing to Pydantic AI (pydantic-ai package, AnthropicModel). ClaudeProvider in src/conductor/providers/claude.py implements the AgentProvider interface while delegating execution details to internal helpers in src/conductor/providers/pydanticai/. There are no user-facing breaking changes. Workflow YAML syntax, runtime.provider: claude configuration, provider contracts, and CLI commands remain completely unchanged. The transition to Pydantic AI includes the following…
About CLAUDE.md
What is CLAUDE.md?
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
Where does it go?
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
What should it contain?
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.md or AGENTS.md?
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.