repo-development
Modify the ADK Python repository itself with repo-specific setup, style, testing, docs, samples, schema generation, and review conventions.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- repo-development
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Modify the ADK Python repository itself with repo-specific setup, style, testing, docs, samples, schema generation, and review conventions.
- disable-model-invocation
- true
02 资源文件 4
- 参考文档references/development-and-testing.md11.5 KB
- 参考文档references/review-checklist.md6.7 KB
- 参考文档references/troubleshooting.md8.8 KB
- 脚本scripts/select_adk_tests.py16.3 KB
03 SKILL.md 正文
ADK Python Repo Development
Use this sub-skill when the user asks to change, review, or validate the ADK Python repository itself. It is for maintainer workflows, not for using ADK as an application library.
Route Common Requests
- Set up or validate a contributor environment: use Development and Testing for
uv, extras, formatter, linter, type-check, pytest, and tox commands. - Select focused tests for a code change: use select_adk_tests.py, which maps changed files or capability names to likely pytest targets and non-test checks without running anything.
- Modify source, public APIs, configs, docs, or samples: use Development and Testing for style rules, public API conventions, generated schema checks, and docs/sample update policy.
- Review local changes before handoff: use Review Checklist for architecture, style, tests, docs, samples, and safety gates.
- Debug repo-maintenance failures: use Troubleshooting for formatter/import issues, optional dependencies, flaky or credentialed tests, stale schemas, and unsafe git operations.
Boundaries
- Use this sub-skill for repository maintenance: files under
src/google/adk/,tests/,docs/,contributing/samples/, packaging/configuration, and contributor scripts. - Route library usage questions about constructing
Agent/LlmAgentto the agent-construction sub-skill. - Route ADK
Workflowgraph design and runtime behavior to workflow-orchestration. - Route tools, toolsets, auth, MCP/OpenAPI/Google/cloud integrations to tools-and-integrations.
- Route
Runner,App, sessions, memory, artifacts, plugins, telemetry, and code execution services to runtime-services. - Route CLI app discovery, YAML config,
adk run,adk web,adk api_server, deploy, and eval command usage to cli-configuration-deployment unless the task is changing the CLI implementation itself. - Route eval/test JSON design and debugging an ADK app run to evaluation-debugging unless the task is changing repo test infrastructure.
Maintainer Defaults
- Package facts:
google-adk2.3.0, import rootgoogle.adk, Python 3.10+. - Base installation may omit optional extras such as
db,extensions,mcp, and cloud integrations; missing optional imports are expected failure modes, not automatic blockers. - Prefer focused validation first, then broaden only when the changed surface or risk justifies it.
- Do not perform destructive git operations, publish packages, run credentialed remote/cloud tests, or rewrite broad areas without explicit user approval.
04 同仓库更多 Skills
passnet-triton-opt
Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.
passnet-solve
END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.
passnet-skill
PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.
passnet-pattern-fusion
Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.
passnet-orchestrate
Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.
passnet-feedback
Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).