k8s-test-infra
NVIDIA/k8s-test-infra/CLAUDE.md
Read ./AGENTS.md file.
How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.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.
NVIDIA/k8s-test-infra/CLAUDE.md
Read ./AGENTS.md file.
NVIDIA/xr-ai/CLAUDE.md
Follow the shared engineering guidance and routing in AGENTS.md, plus the nearest package or application README. For repository contributions, follow Contributing and conventions. Run relevant tests and Ruff before committing, and sign repository commits with git commit -s. After pushing an upstream PR, check gh pr checks; the change is not complete until required CI is green. Never launch models ad hoc (bare vllm serve, from_pretrained scripts). Start servers only through the repo entry points (sample orchestrators and the services/…
anthropics/single-agent-loop-kit/CLAUDE.md
This is the single agent loop kit: a Claude Code skill (skill/), four scripts, a reference loop, and a runnable example. README.md has the map. Follow examples/support-desk/RUNBOOK.md. During the audit, do not open examples/support-desk/ANSWER-KEY.md, examples/support-desk/single_loop/ or examples/support-desk/example-run/. They give away the verdicts the audit is supposed to reach from the code and traces. The skill belongs in their agent repo, not here. Point them at "Use it on your own system" in README.md. - Run python3 -m unittest discover tests…
facebookresearch/visual_inertial_bundle_adjustment/CLAUDE.md
No summary in the file. Open it to read it.
microsoft/agentic-agile-template/platform-adapters/agent-tools/claude/CLAUDE.md
This is an agent context file. It provides AI coding agents (GitHub Copilot, Claude, Cursor, or any LLM-based development tool) with the information they need to work effectively in your repository. Think of it as onboarding documentation, but optimized for agents: structured, explicit, and maintained alongside your code. WHY DOES IT MATTER? Without an agent context file, every agent session starts from zero. The agent must infer your conventions, guess your architecture, and hope its assumptions match yours. With a…
vercel-labs/phase/CLAUDE.md
See AGENTS.md for repository instructions and its package-specific instruction pointers.
JetBrains/defender_performance_tool/src/DefenderPerformanceTool/CLAUDE.md
There are no automated tests; verification is manual by running the app and observing ETW events during a Defender scan. A WPF application with three components wired together in Program.cs: EtwListener.cs — subscribes to the Microsoft-Antimalware-Engine ETW provider on a background thread. Correlates Start/Stop Streamscanrequest events using ActivityID (Guid), calculates scan duration, and emits EventInfo records via a reactive Subject<EventInfo>. Plotter.cs — subscribes to the IObservable<EventInfo> stream. Accumulates durations per-process per-second in a ConcurrentDictionary, then a DispatcherTimer fires every second…
Shopify/worldwide/CLAUDE.md
The worldwide gem provides internationalization and localization APIs for Ruby, including address validation/formatting, currency handling, date/time formatting, region/country data, phone validation, and more. It also includes a TypeScript/JavaScript package at…
cloudflare/vibesdk-templates/CLAUDE.md
This repository contains the Cloudflare VibeSDK template catalog - a dynamic template generation system for AI-assisted application development. Templates are not stored as complete projects but are dynamically generated from…
NVIDIA/srt-slurm/CLAUDE.md
Development guide for working on this codebase. For every Slurm run you launch or analyze, write and maintain <run_dir>/logs/perf-analysis.md in the actual job's logs directory on the cluster. Read and…
databricks/zerobus-sdk/CLAUDE.md
For agents working on the SDK. To use an SDK, see README.md and the SDK's own README.md. The one client-code rule is in the last section. One Rust core (rust/sdk/), exposed through several bindings; two SDKs do not use it. Run from the SDK directory.
databricks/zerobus-sdk/cpp/CLAUDE.md
C++ wrapper around the Rust core via the C FFI library (rust/ffi/zerobus.h). Like Go, it consumes the C FFI directly; unlike Go there is no GC, so resource management is RAII and errors are reported by throwing zerobus::ZerobusException. The public headers never include zerobus.h; they hold opaque, forward-declared FFI handles. Only src/ includes the generated C header, keeping the C FFI an implementation detail of the SDK. Run from cpp/: - make build — Configure (CMake) and build the SDK,…
databricks/zerobus-sdk/go/CLAUDE.md
Go wrapper around the Rust core via cgo and the C FFI library. When writing or reviewing client/example code, follow the idiomatic async flow. IngestRecordOffset() and IngestRecordsOffset() (on ZerobusStream), as well as IngestBatch() (on ZerobusArrowStream), return as soon as the record is queued; the SDK sends it and tracks its acknowledgment in the background. - Ingest in a loop, then call Flush() to confirm durability — once for a bounded batch, or periodically for a long-running stream. - Acks are…
databricks/zerobus-sdk/java/CLAUDE.md
Java wrapper around the Rust core via JNI. Ingestion is asynchronous. ingestRecordOffset() (and ingestRecordsOffset()) return as soon as the record/batch is queued — the SDK sends it and tracks its acknowledgment in the background. - Idiomatic flow: ingest in a loop, then call flush() once (after a bounded batch, or periodically for a long-running stream) to confirm durability. Equivalently, call waitForOffset() on the last returned offset — acks are ordered, so confirming the last offset confirms every prior record. -…
databricks/zerobus-sdk/python/CLAUDE.md
Python wrapper around the Rust core via PyO3 and maturin. When writing or reviewing client code that uses this SDK, follow the cross-SDK performance flow from the root CLAUDE.md: - Idiomatic flow: ingest in a loop with ingest_record_offset(), or prefer ingest_records_offset() for bulk data, then call flush() once to confirm durability. ingest_record_offset() returns as soon as the record is queued; the SDK sends it and tracks its acknowledgment in the background. Do not feature ingest_record_nowait() in examples: it spawns a…
databricks/zerobus-sdk/rust/CLAUDE.md
The core SDK; every Rust-backed SDK binds it: Root CLAUDE.md has the shared rules; this is the Rust-core delta. The JSON/protobuf engine is in sdk/src/stream/grpc/; Arrow Flight is in sdk/src/stream/arrow/ (feature arrow-flight). From rust/: make build / build-release, build-ffi (rebuilds zerobus.h), build-jni, lint, fmt, check. Not obvious: A public-surface change reaches: pub items → Rust consumers; FFI signatures/structs → Go/C++/.NET; JNI methods → Java; PyO3 (python/rust/src/lib.rs) and NAPI (typescript/src/lib.rs) wrap the SDK directly, so refactors can break them via changed…
databricks/zerobus-sdk/typescript/CLAUDE.md
Node.js wrapper around the Rust core via NAPI-RS. When writing or reviewing client code that uses this SDK, prefer the idiomatic ingest-then-flush flow. ingestRecordOffset() (and ingestRecordsOffset()) resolves as soon as the record is queued — the SDK sends it and tracks its acknowledgment in the background. Ingest in a loop, then call flush() once to confirm durability (once for a bounded batch, or periodically for a long-running stream). Each ingest returns the record's offset, and waitForOffset(offset) confirms a specific record…
microsoft/Agents-for-js/CLAUDE.md
This is the Microsoft 365 Agents SDK for JavaScript/TypeScript (formerly known as BotFramework SDK v5). It provides a comprehensive framework for building enterprise-grade conversational agents that work across M365, Teams, Copilot Studio, Webchat, and other platforms. The repository is a monorepo using npm workspaces with multiple interconnected packages. Sample agents used for internal testing ("samples"). For learning examples, refer to https://github.com/Microsoft/Agents. Build Process: Uses TypeScript project references defined in tsconfig.build.json. Each package builds independently but respects dependencies through references. Test…
aws/modern-data-architecture-accelerator/CLAUDE.md
See developer-code-review Operational guidance for commits, issues, MRs, reviews, and repo workflows See developer-operational-guidance Non-optional working requirements for all tasks See developer-working-requirements
modelcontextprotocol/example-remote-server/CLAUDE.md
Always build, lint, and test your changes.
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.