菜单
Analyze a codebase before changing it, infer local patterns, and choose focused implementation and verification steps.
Elixir development guidance for creating, modifying, reviewing, testing, and debugging Elixir and Phoenix applications. Use when tasks involve .ex or .exs files, mix.exs, Phoenix, LiveView, Ecto, OTP supervision trees, GenServer, ExUnit, mix format, mix test, migrations, schemas, contexts, or Elixir dependency management.
Go development guidance for creating, modifying, reviewing, testing, and debugging Go services, libraries, CLIs, and tooling. Use when tasks involve .go files, go.mod, go.work, goroutines, channels, contexts, HTTP handlers, database code, table tests, gofmt, go test, or Go module management.
Node.js and JavaScript/TypeScript development guidance for creating, modifying, reviewing, testing, packaging, and debugging backend, CLI, tooling, and full-stack code. Use when tasks involve package.json, npm, pnpm, yarn, bun, .js, .mjs, .cjs, .ts, tsconfig, Express, NestJS, Next.js API routes, Vitest, Jest, ESLint, or Node dependency management.
Python development guidance for creating, modifying, reviewing, testing, packaging, and debugging Python code. Use when tasks involve .py files, pyproject.toml, requirements files, pytest, ruff, mypy, FastAPI, Django, Flask, CLIs, scripts, type hints, virtual environments, or Python dependency management.
Rust development guidance for creating, modifying, reviewing, testing, packaging, and debugging Rust crates, CLIs, services, and libraries. Use when tasks involve .rs files, Cargo.toml, Cargo.lock, workspaces, ownership, lifetimes, traits, async Rust, tokio, serde, cargo fmt, cargo test, cargo clippy, or Rust dependency management.
Planning discipline for long-running Deep Agent tasks with delegation, task tracking, validation, and concise delivery.
| name | codebase-analysis |
| description | Analyze a codebase before changing it, infer local patterns, and choose focused implementation and verification steps. |
| metadata | {"version":"0.2.0","tags":["code","analysis","testing"]} |
Use this skill when the task involves reading, changing, reviewing, or explaining code.
Build confidence from local evidence before changing code. This skill favors small, repeatable investigation steps that work with both small local models and larger hosted models.
When the selected model is small or local:
Focus review comments on:
Keep the final answer compact:
