agent-framework / ollama
microsoft/agent-framework/python/packages/ollama/AGENTS.md
Integration with Ollama for local LLM inference.
How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.
microsoft/agent-framework/python/packages/ollama/AGENTS.md
Integration with Ollama for local LLM inference.
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 functionapprovalrequest is server-issued and must not be replayed inline, while the new hosted functionapprovalresponse is serialized as mcpapprovalresponse 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.…
microsoft/agent-framework/python/packages/purview/AGENTS.md
Integration with Microsoft Purview for data governance and policy enforcement.
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. - RedisContextProvider - Context provider with Redis-backed retrieval - RedisSettings - TypedDict connection settings for vector stores, resolved with core loadsettings from explicit URL overrides, an optional .env file,…
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…
siteboon/claudecodeui/AGENTS.md
For every task that creates, modifies, refactors, or reviews backend code under server/, load and follow $backend-module-standards from .agents/skills/backend-module-standards/SKILL.md. Apply it only to backend code; do not impose those architecture…
XiaomiMiMo/MiMo-Code/AGENTS.md
Development focuses on the TUI (packages/opencode/src/cli/cmd/tui/) and the engine core it runs on. The Web, App and Desktop surfaces are not currently maintained. In practice this means an operation should default to checking the TUI path first. Reduce total variable count by inlining when a value is only used once. Avoid unnecessary destructuring. Use dot notation to preserve context. Prefer const over let. Use ternaries or early returns instead of reassignment. Avoid else statements. Prefer early returns. Use snake_case for…
huggingface/speech-to-speech/AGENTS.md
PyPI publishing is handled by GitHub Actions in .github/workflows/publish.yml. The workflow runs on pushed tags that match v*, builds the package with uv build, checks the artifacts with twine check --strict, and publishes through the configured pypi environment. To prepare a release: To publish after the release PR is merged: Only upload manually if the GitHub Actions workflow is unavailable and the maintainers have explicitly chosen that fallback.
sirmalloc/ccstatusline/AGENTS.md
ccstatusline is a customizable status line formatter for Claude Code CLI that displays model info, git branch, token usage, and other metrics. It functions as both: 1. A piped command processor for Claude Code status lines 2. An interactive TUI configuration tool when run without input The project has dual runtime compatibility - works with both Bun and Node.js: Widget Interface: All widgets must implement: - getDefaultColor(): Default color for the widget - getDescription(): Description shown in TUI - getDisplayName():…
aws/aws-cdk/AGENTS.md
Contributor-focused guide for AI agents working on the AWS CDK codebase. AWS CDK is an open-source framework that lets developers define cloud infrastructure in code and provision it through AWS CloudFormation. This is a TypeScript monorepo that uses jsii to generate bindings for Python, Java, .NET, and Go. Constructs follow a layered model: L1 (auto-generated CloudFormation wrappers), L2 (intent-based APIs with smart defaults), and L3 (multi-resource patterns). See CONTRIBUTING.md for the full contributor guide. You are a CDK contributor. You…
google/oss-fuzz/AGENTS.md
No summary in the file. Open it to read it.
cloudflare/quiche/AGENTS.md
Generated: 2026-02-20 Commit: 89d1850f Branch: master Cloudflare's QUIC and HTTP/3 implementation in Rust. Workspace of 11 crates: core quiche protocol library, tokio-quiche async integration, CLI tools (apps, h3i), logging/analysis (qlog, qlog-dancer, netlog), and supporting primitives (octets, buffer-pool, datagram-socket, task-killswitch).
cloudflare/quiche/datagram-socket/AGENTS.md
UDP socket abstraction layer shared across the quiche workspace. Two main concerns: DgramBuffer (a headroom-aware zero-copy byte buffer for datagrams) and DatagramSocket* traits (uniform async send/recv over UdpSocket/UnixDatagram with Linux recvmmsg/sendmmsg batching). Wraps a Vec<u8> with a pos cursor dividing it into [headroom: 0..pos][payload: pos..]. as_slice() / len() expose only the payload. BufMut writes append to the tail, independent of pos. Fast path: allocate with withcapacityandheadroom, fill payload via BufMut, call tryadd_prefix to slide pos left and write the prefix…
cloudflare/quiche/h3i/AGENTS.md
Low-level HTTP/3 testing client. Sends arbitrary/malformed H3 frames to probe server RFC compliance. Both library (lib.rs) and CLI binary (main.rs). Used programmatically as test driver in tokio-quiche integration tests.
cloudflare/quiche/qlog/AGENTS.md
qlog data model for QUIC and HTTP/3 per IETF drafts (draft-ietf-quic-qlog-main-schema, draft-ietf-quic-qlog-quic-events, draft-ietf-quic-qlog-h3-events). Pure data types + serde serialization; no IO (deferred to consumers).
cloudflare/quiche/qlog-dancer/AGENTS.md
Visualization/analysis tool for qlog and Chrome netlog files. Parses logs into a Datastore, extracts time-series into SeriesStore, renders PNG charts (native) or canvas plots (wasm). Outputs HTML/text reports with tabled summaries. Dual-target: native CLI binary (main.rs) + wasm-bindgen web UI (web.rs, #[cfg(target_arch = "wasm32")]). crate-type = ["lib", "cdylib"] -- cdylib is for wasm.
cloudflare/quiche/quiche/AGENTS.md
Low-level QUIC transport and HTTP/3 in Rust. App provides IO/timers; this crate handles protocol state. Also exposes C FFI via staticlib/cdylib. BoringSSL itself is built by boring-sys (no in-tree submodule).
cloudflare/quiche/quiche/src/h3/AGENTS.md
HTTP/3 wire protocol over QUIC. Connection manages H3 state (streams, QPACK, SETTINGS, GOAWAY) on top of quiche::Connection<F>. Event-driven: caller loops poll() → Event. Own Error/Result types, separate from quiche::Error.
cloudflare/quiche/quiche/src/recovery/AGENTS.md
QUIC loss detection and congestion control per RFC 9002. Two parallel CC implementations coexist: congestion/ (legacy Reno/CUBIC) and gcongestion/ (next-gen BBR2 ported from google/quiche). Recovery enum dispatches between LegacyRecovery and GRecovery via enum_dispatch over the RecoveryOps trait (40+ methods). - Two Acked structs at congestion/recovery.rs:1079 and gcongestion/mod.rs:49. NOT unified. Do NOT create a third. - FIXME stubs: Some RecoveryOps methods only apply to one impl. Both sides have // FIXME only used by {congestion,gcongestion} stubs that return defaults. Do not…
cloudflare/quiche/tokio-quiche/AGENTS.md
Async tokio wrapper for quiche. Spawns per-connection IO worker tasks driven by an ApplicationOverQuic trait. Ships a ready-made H3Driver for HTTP/3. Uses foundations for structured logging (slog), telemetry, and settings.
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.