MicrosoftAGENTS.mdcates / level-3
microsoft/cates/examples/level-3/AGENTS.md
No summary in the file. Open it to read it.
AGENTS.md examples from real projects
How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Worked for most · soon
MicrosoftAGENTS.mdMicrosoftAGENTS.mdcopilot-brag-sheet
microsoft/copilot-brag-sheet/AGENTS.md
README for AI coding agents. If you're a human looking for setup or usage instructions, read README.md instead. This file is the dedicated place where Claude Code, Copilot CLI, Codex, Cursor, Jules, Amp, and friends find the context they need to be productive on this repo without re-deriving it from grep. Format: agents.md — used by 60k+ open-source projects and stewarded by the Agentic AI Foundation under the Linux Foundation. The closest AGENTS.md to the file you're editing wins; explicit…
MicrosoftAGENTS.mdidentity-spiffe
microsoft/identity-spiffe/AGENTS.md
Canonical contributor and AI-agent guidance. Durable architecture and runbooks live in docs/. Before changing authentication, deployment, or authorization code, read: Common flows: Use python3 scripts/test_agents.py to validate the live enforcement matrix after deployment.
MicrosoftAGENTS.mdidentity-spiffe / infra
microsoft/identity-spiffe/infra/AGENTS.md
All Azure infrastructure-as-code for the Identity Research for Agent Management Using SPIFFE. Deployed via azd provision (Step 1 of deploy.sh). Never edit Container App container specs in Bicep — deploy.sh manages the full container configuration to avoid the multi-container reset bug (Learning #3).
MicrosoftAGENTS.mdidentity-spiffe / scripts
microsoft/identity-spiffe/scripts/AGENTS.md
Deployment, testing, and identity management scripts. When only the cloud portal UX/auth surface changed, prefer ./deploy.sh --portal-only. That path rebuilds and updates isp-portal + securityportal-mock without touching agent sidecars or forcing SPIRE re-attestation. Read alongside: The enforcement matrix validation script. Runs all 22 scenarios and reports pass/fail. Transport Layer (mTLS): 1. BudgetReport → BudgetBackend: ✅ connection succeeds 2. EmployeeMenus → BudgetBackend: ❌ connection refused (mTLS rejection) 3. BudgetApproval → BudgetBackend: ✅ connection succeeds Application Layer (RBAC): 4. BudgetReport → GET…
MicrosoftAGENTS.mdidentity-spiffe / tests
microsoft/identity-spiffe/tests/AGENTS.md
Read README.md and the relevant suite README before running tests. This file is the operational contract for an LLM asked to validate a change. 1. Identify the changed components and requested profile. Inventory with python3 tests/run.py --list; choose suites deliberately. 2. Check prerequisites without dumping credentials, browser state, or .azure configuration. Do not read other agents' recovery/session artifacts. 3. Run harness self-tests when changing the harness. Run local browser/protocol and connected E2E coverage for relevant application behavior. After setup, tests/.venv/bin/python…
MicrosoftAGENTS.mdagora-workbench
microsoft/agora-workbench/AGENTS.md
The project uses uv for dependency management with pyproject.toml at the repo root. Python ≥3.11 is required. Always run commands from the repo root. Use uv run to run any script (e.g. uv run python src/script.py) and uv add to add packages. The virtual environment is managed automatically — do not manually activate .venv. src/ is on the Python path. All imports use the agora_workbench namespace: Do not use bare package names — from code_execution import ... is wrong. BaseMCPServer…
MicrosoftAGENTS.mdazure-devops-copilot-plugin / azure-devops
microsoft/azure-devops-copilot-plugin/extensions/azure-devops/AGENTS.md
No summary in the file. Open it to read it.
MicrosoftAGENTS.mdCopilot-Managed-Runtime / github-actions
microsoft/Copilot-Managed-Runtime/github-actions/AGENTS.md
This directory holds the install-ms-cli, ms-app-pack, and ms-app-deploy GitHub Actions for building and deploying apps on Copilot Managed Runtime from CI. If you're an agent (GitHub Copilot CLI, Claude Code, or any other) helping with this component, the reusable skills below are plain Markdown runbooks — read the relevant one and follow its steps. - Build the actions: cd github-actions && npm install && npm run build (compiles src/ and bundles each action into dist/actions/<name>/index.js). - dist/ is committed and…
MicrosoftAGENTS.mdCopilot-Managed-Runtime / copilot-managed-runtime
microsoft/Copilot-Managed-Runtime/plugins/copilot-managed-runtime/AGENTS.md
This file provides instructions for assistants working on the Copilot Managed Runtime plugin itself (not for end-users of the plugin). The Copilot Managed Runtime plugin wraps @microsoft/cmr-cli (binary ms) to scaffold, develop, and ship apps on Copilot Managed Runtime end-to-end. It provides skills for: This plugin uses a memory bank (memory-bank.md) to persist state across sessions.
MicrosoftAGENTS.mdcrap4csharp
microsoft/crap4csharp/AGENTS.md
Refer .github/copilot-instructions.md.
MicrosoftAGENTS.mdfrontier-agentic-devops-rvas
microsoft/frontier-agentic-devops-rvas/AGENTS.md
Remember the law of diminishing returns to decide if something should fit or not here. Before drafting, editing, or returning repository prose, invoke /humanize-content. This includes technical documentation, session kits, runbooks, Markdown, slide content, and user-facing explanations. If the skill is not loaded, read and follow its instructions directly. Use the skill to simplify wording and remove AI-patterns from technical content. Use the clear-thinker voice unless the user requests another voice. Before finishing, complete both checks: 1. Apply the prose…
MicrosoftAGENTS.mdAKS-Lab-GitHubCopilot
microsoft/AKS-Lab-GitHubCopilot/AGENTS.md
Inspired by microsoft/agent-framework/python/AGENTS.md. Read this before you write any code in this repo — whether you are a human, GitHub Copilot, or the Copilot Coding Agent. This repository implements ZavaShop, a multi-agent retail supply-chain demo. It runs on AKS (long-lived orchestrator) and Azure Container Apps (event-driven specialist agents). Agents are built with Microsoft Agent Framework (MAF) and the GitHub Copilot SDK Python provider. Treat every change as production-style: typed, tested, observable. This repo is delivered through GitHub Copilot Custom Agents.…
MicrosoftAGENTS.mdamplifier-agent
microsoft/amplifier-agent/AGENTS.md
amplifier-agent is the Amplifier agent as a library: Python and TypeScript bindings and an HTTP face over one replaceable engine. Pages about internals (docs/development/, docs/DEVELOPMENT.md) change freely but stay accurate. 3. Evaluations. Is the behavior covered by a task in evaluations/? Add one when it is not, and keep evaluations/README.md current when the harness changes. 4. Tests. Add unit or integration tests where they help implementation or pin precise internal behavior an evaluation cannot observe. - For a new feature,…
MicrosoftAGENTS.mdamplifier-agent / workspace
microsoft/amplifier-agent/evaluations/tasks/core/agents_md/workspace/AGENTS.md
End every Python file with the comment line # end of module.
MicrosoftAGENTS.mdamplifier-agent / repo
microsoft/amplifier-agent/evaluations/tasks/core/agents_md/workspace/repo/AGENTS.md
No summary in the file. Open it to read it.
MicrosoftAGENTS.mdamplifier-agent / service
microsoft/amplifier-agent/evaluations/tasks/core/agents_md/workspace/repo/service/AGENTS.md
Python files in this directory start with # owner: team-osprey in place of the repository's owner line.
MicrosoftAGENTS.mdCopilotAnalyticsLabs
microsoft/CopilotAnalyticsLabs/AGENTS.md
Labs is opened from inside Microsoft's in-product analytics portal and must look like part of it. Every UI change follows the portal design language. 1. docs/portal-design-guidelines.md: tokens, type roles, spacing,…
MicrosoftAGENTS.mdamplifier-bundle-digital-twin-universe
microsoft/amplifier-bundle-digital-twin-universe/AGENTS.md
Conventions for AI coding agents (Amplifier, Claude Code, Cursor, etc.) and human contributors using them. Read this before making changes. @README.md @docs/development.md @docs/api-reference.md @docs/profiles.md @docs/amplifier-user-sim-launch-flow.dot @agents/dtu-profile-builder.md @profiles/tests/amplifier-user-sim.yaml @profiles/amplifier/amplifier-chat.yaml CLI and library for spawning and managing Digital Twin Universe (DTU) environments. Wraps the incus CLI with profile-driven provisioning, file push/pull (recursive), exec, and lifecycle management. The DTU is how the broader ecosystem gets an isolated, realistic environment for integration verification. Run these before opening a PR. The reviewer expects evidence in…
MicrosoftAGENTS.mdamplifier-bundle-recipes
microsoft/amplifier-bundle-recipes/AGENTS.md
Conventions for any agent (or human) changing amplifier-bundle-recipes. Run all four before opening a PR: Never re-record the legacy-compat baselines to make a change pass. They pin the serialized ToolResult payload of execute / resume / approve for legacy recipes. If your change needs to add something to a result, attach it beside the payload (object.__setattr__(result, ...), as runner_adapter.label_execution_mode, shutdown.attach_shutdown_warning and engine_provenance.label_engine_provenance all do), not inside output. Both are generated artifacts, not hand-maintained ones. They are produced from the repo…
About AGENTS.md
What is AGENTS.md?
An open format for instructions to coding agents, read by Codex, Cursor and others. Think of it as a README written for agents.
Where does it go?
At the repository root, with more specific files in subdirectories. Agents read the one closest to the file they're editing.
What should it contain?
Setup and test commands, code style, and the rules a new contributor would need to know.
Does Claude Code read it?
Claude Code reads CLAUDE.md. A one-line CLAUDE.md that points at AGENTS.md covers both.