agent-framework / copilotstudio
microsoft/agent-framework/python/packages/copilotstudio/AGENTS.md
Integration with Microsoft Copilot Studio agents.
Steady10 mentions in 2 weeks
No other project uses it yet
How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.
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.
microsoft/agent-framework/python/packages/copilotstudio/AGENTS.md
Integration with Microsoft Copilot Studio agents.
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/core/AGENTS.md
The foundation package containing all core abstractions, types, and built-in OpenAI/Azure OpenAI support. - agent_framework.__init__ uses lazy module-level __getattr__ for most public exports to keep cold import agent_framework lightweight. - Keep _LAZY_MODULE_EXPORTS, _LAZY_EXPORTS, the explicit runtime __all__, and __init__.pyi synchronized whenever adding, removing, or moving a root public export. - Runtime __all__ is still required for from agent_framework import *; the .pyi file is for type checkers and editors and does not replace runtime exports. - Public deprecation behavior for…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/declarative/AGENTS.md
YAML/JSON-based declarative agent and workflow definitions. Both state classes validate object attribute names but leave dictionary keys unchanged. DefaultHttpRequestHandler(max_response_body_bytes=...) defaults to a 10 MiB wire and decoded-stage body limit, including redirect and authentication responses. It supports identity, gzip, and deflate, rejects unsupported or incomplete encodings, and closes response streams without closing caller-owned clients. HTTPX is constrained to the 0.27 and 0.28 series; focused CI checks both series because response caching and default-header integration use its internals. DefaultMCPToolHandler creates a fresh…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/devui/AGENTS.md
Interactive developer UI for testing and debugging agents and workflows. Text content carrying additional_properties["core"]["model_output_kind"] == "refusal" is mapped to native Responses refusal parts and events; the legacy flat key remains readable. OpenAI input metadata uses the openai namespace while replay-critical writers retain documented legacy keys. Mapper aggregation, live rendering, and recovery state retain text/refusal boundaries by item and content index. DevUI is a development-only sample app, not a production hosting surface. Authentication is enabled by default. Unauthenticated mode is allowed…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/foundry_local/AGENTS.md
Integration with Microsoft Foundry Local for local model inference. Uses foundry_local_sdk (SDK 2.3.1+, below 3.0) on Python 3.11-3.14. Clients share FoundryLocalManager.instance and use its built-in web service with the existing OpenAI transport. Model lifecycle operations live on catalog model objects. The adapter defaults to the public catalog and /v1. Use catalog_type="local" or foundry_local_sdk.catalog.CatalogType.LOCAL for SDK-registered BYOM models: both lookup and HTTP inference use the local catalog and /catalogs/local/v1. Catalogs never fall back to each other; per-request model overrides remain in…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/gemini/AGENTS.md
Integration with Google's Gemini Developer API and Enterprise (Vertex AI) via the google-genai SDK. The shared _sdk_client.create_genai_client resolves authentication, backend mode, and service URL for chat and embeddings. GeminiEmbeddingClient supports only gemini-embedding-2 and gemini-embedding-2-preview, requiring per-call task instructions for text strings. Multimodal Google SDK Content or media Part inputs receive no task prefix, even with mixed text-and-media parts. Text-only SDK content is rejected so callers cannot bypass the task requirement. Enterprise accepts one content per request; the client splits batches…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/github_copilot/AGENTS.md
Integration with GitHub Copilot extensions. _build_session_kwargs forwards options to the SDK verbatim except for a few keys that get an explicit default: on_permission_request (deny-all) and enable_file_hooks (False, so a session behaves the same way in every working directory). Callers opt in through default_options or per-run options. Keep such defaults to options the working directory controls: options that only shape prompt context (for example enable_host_git_operations) are deliberately left alone. When the agent applies the enable_file_hooks default and the working directory actually…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/hosting-a2a/AGENTS.md
Conversion and native card-generation helpers for hosting Agent Framework agents and workflows through an application-owned A2A server. - a2a_to_run(message, *, stream=False) converts an A2A Message to AgentRunArgs; pass input_modes for optional advertised-mode validation. - a2a_from_run(result) converts an Agent Framework response, message, or streaming update to A2A Part values; pass output_modes for optional advertised-mode validation. - await AgentA2AAdapter(target, ...).get_card() creates a native AgentCard from an agent or AgentState, target metadata, and explicit A2A discovery policy. The adapter also re-exposes a2a_to_run(...) and…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/hosting-mcp/AGENTS.md
Side-effect-free adapters and conversion helpers for hosting Agent Framework agents and workflows through the native MCP SDK. - AgentMCPTool(target, ...) generates one native MCP Tool from an agent, converts and executes calls, and optionally persists sessions through an existing AgentState. - WorkflowMCPTool(target, ...) derives one native MCP Tool from a workflow's single start-executor input type and converts completed workflow outputs. - mcp_to_run(arguments, *, argument_name="task", chat_option_arguments=()) converts MCP tool arguments to AgentRunArgs and copies only explicitly selected arguments into chat options.…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/lab/AGENTS.md
Experimental packages for cutting-edge features including benchmarking, reinforcement learning, and research initiatives. This package contains experimental sub-packages: Lab packages are experimental and may change frequently. They are not included in the standard agent-framework[all] installation. Lab is a standalone uv project, excluded from the root python workspace so its experimental dependencies do not constrain released provider packages. It resolves released Core and provider distributions by default. Run dependency and validation commands from python/packages/lab. Lab code must use APIs available in published…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/mem0/AGENTS.md
Integration with Mem0 for agent memory management. - application_id / agent_id / user_id are the storage scope stamped on written memories. - search_application_id / search_agent_id / search_user_id are the retrieval scope. - Retrieval scope never defaults to the storage scope. With no search_* value set, before_run retrieves nothing and logs a warning. This prevents memories written under a shared agent_id from being retrieved for every user of that agent. Mem0 telemetry is disabled by default. Set MEM0_TELEMETRY=true to enable.
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/mistral/AGENTS.md
Integration with Mistral AI for chat completions and embedding generation.
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/monty/AGENTS.md
Monty-backed CodeAct integrations for the Microsoft Agent Framework. [!NOTE] Beta package. Included in agent-framework[all] and available through the lazy-loading namespace agent_framework.monty. - MontyCodeActProvider — ContextProvider that injects a run-scoped execute_code tool plus dynamic CodeAct instructions. Mirrors the HyperlightCodeActProvider API for the parts that apply to a non-sandboxed Python interpreter. - MontyExecuteCodeTool — FunctionTool that wraps the Monty interpreter. Use directly for mixed-tool agents or manual static wiring. Mirrors HyperlightExecuteCodeTool. MontyCodeActProvider and MontyExecuteCodeTool both accept: - tools — host tool callables…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/nvidia/AGENTS.md
Integration with NVIDIA NIM for embedding generation. - NVIDIA NIM exposes an OpenAI-compatible endpoint, so this package builds on RawOpenAIEmbeddingClient from agent-framework-openai rather than reimplementing the wire format. This mirrors agent-framework-foundry-local, which builds on the OpenAI Chat Completions client for the same reason. - The default endpoint is https://integrate.api.nvidia.com/v1. base_url overrides it, which is the hook for a self-hosted NIM container. - Model availability is per-account. An unentitled model returns 404 ... Not found for account, and a retired model…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/ollama/AGENTS.md
Integration with Ollama for local LLM inference.
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/openai/AGENTS.md
OpenAI integration package for Agent Framework. Contains OpenAI Responses API and Chat Completions API clients. All clients follow the Raw + Full-Featured pattern (e.g., RawOpenAIChatClient + OpenAIChatClient). For Responses API continuation with service-side storage, a prior hosted function_approval_request is server-issued and must not be replayed inline, while the new hosted function_approval_response is serialized as mcp_approval_response so the user's approved or rejected decision reaches the service. Local FunctionTool approval controls are resolved in-process and must not be serialized as MCP items.…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/purview/AGENTS.md
Integration with Microsoft Purview for data governance and policy enforcement.
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/redis/AGENTS.md
Redis-based storage for agent threads and context. - RedisHistoryProvider - Persistent chat history provider using scoped Redis keys. Scoped mode requires an application ID and non-empty session ID, and can additionally isolate tenants and agents. Use explicit legacy mode only while deliberately migrating historical keys; scoped mode never accesses them automatically. It can construct an owned client from a URL or credential provider, or borrow a standalone async Redis client configured with decode_responses=True; borrowed clients remain caller-owned. - RedisContextProvider -…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/packages/typesafe/AGENTS.md
Integration with TypeSafe AI System One models, including Jev. - TypeSafeChatClient - Adds function invocation, middleware, and telemetry to the TypeSafe transport. - RawTypeSafeChatClient - Provider transport that can emit constrained function calls but does not execute them; it has no function-invocation, middleware, or telemetry layers. - TypeSafeChatOptions - Uses response_format for the required TypeSafe Questions mapping. - Calls always return structured SystemOneResponse data. This provider does not generate free-form chat text. - Every call requires response_format to contain at…
Steady10 mentions in 2 weeks
No other project uses it yet
microsoft/agent-framework/python/samples/AGENTS.md
This file documents the structure and conventions of the Python samples so that agents (AI or human) can maintain them without rediscovering decisions. Durable Task and Azure Functions samples are maintained in the Durable Agent Framework extension. 1. Progressive complexity: Sections 01→05 build from "hello world" to production. Within 01-get-started, files are numbered 01–07 and each step adds exactly one concept. 3. Workflows preserved: 03-workflows/ keeps the upstream folder names and file names intact. Do not rename or restructure workflow…
Steady10 mentions in 2 weeks
No other project uses it yet
An open format for instructions to coding agents, read by Codex, Cursor and others. Think of it as a README written for agents.
At the repository root, with more specific files in subdirectories. Agents read the one closest to the file they're editing.
Setup and test commands, code style, and the rules a new contributor would need to know.
Claude Code reads CLAUDE.md. A one-line CLAUDE.md that points at AGENTS.md covers both.