squirrel
Full-cycle AI coding skill: plans, builds, tests, lints, fixes bugs, and writes production-grade docs. Auto-detects project state and adapts its 8-phase pipeline.
- License 许可
- Apache-2.0
- 最近更新
- 2026-10-23
- Stars
- ★ 5
- Forks
- 0
01 前言 Front Matter
- name
- squirrel
- risk
- safe
- tags
- [ "development", "testing", "planning", "code-review", "documentation", "ci-cd" ]
- tools
- [ "claude", "cursor", "codex", "antigravity", "gemini", "windsurf", "opencode", "copilot" ]
- author
- flying_squirrel__
- source
- community
- license
- Apache-2.0
- category
- development
- date_added
- 2026-04-29
- description
- Full-cycle AI coding skill: plans, builds, tests, lints, fixes bugs, and writes production-grade docs. Auto-detects project state and adapts its 8-phase pipeline.
- source_repo
- flyingsquirrel0419/squirrel-skill
- source_type
- community
- license_source
- https://github.com/flyingsquirrel0419/squirrel-skill/blob/main/LICENSE
03 SKILL.md 正文
Squirrel — Full-Cycle Software Development Skill
Overview
Squirrel is a full-cycle AI coding skill that works across 9 AI coding agents. It auto-detects project state (greenfield, in-progress, or mature) and adapts its 8-phase engineering pipeline accordingly. Instead of a one-size-fits-all workflow, it figures out where the project actually is and jumps in at exactly the right point.
When to Use This Skill
- Use when starting a new project from scratch (greenfield)
- Use when improving an existing codebase (in-progress or mature)
- Use when fixing bugs, adding features, or refactoring
- Use when adding tests, linting, or CI/CD to a project
- Use when writing production-grade documentation
- Use when the user says "build me", "fix this", "squirrel this project", or any multi-step development task
How It Works
Step 0: Detect Mode
Squirrel classifies the project directory:
| Signal | Mode | Entry Point |
|---|---|---|
| Empty directory | Greenfield | All 8 phases from scratch |
| Source files, no tests/docs | In-Progress | Audit first, then improve |
| Source + tests + CI + README | Mature | Targeted improvements |
| "fix this bug / add feature" | Targeted | Scoped work only |
The 8-Phase Pipeline
- Discover — Understand the project (audit existing code or gather requirements)
- Plan — Concrete task list with dependencies and done-criteria
- Build — Write or modify code (parallel sub-agents when supported)
- Test — Run existing tests, write new ones, 70%+ coverage target
- Bug Hunt — Static analysis + manual review
- Polish — Lint, format, type check, remove dead code
- Document — README + inline docs (update existing, don't overwrite)
- Ship — Final checklist: tests green, no secrets, CI configured
Failure Recovery (3-Strike Rule)
- Strike 1: Fix the specific error. Run tests. Move on.
- Strike 2: Re-read the code. Try a different approach.
- Strike 3: STOP. Revert. Document what failed. Ask the user.
Examples
Example 1: Build a REST API
> build me a REST API for a todo app with TypeScript and Express
Squirrel auto-detects greenfield mode and runs all 8 phases.
Example 2: Fix a bug
> fix this bug in src/auth/login.py
Squirrel enters targeted mode — abbreviated audit, scoped fix, verify.
Example 3: Improve existing project
> squirrel this project — add tests, fix lint errors, write README
Squirrel audits the existing codebase, then applies phases 4-8.
Best Practices
- Respects existing code — matches naming conventions, test framework, import style, and architecture
- Reads 2-3 similar files before writing a new one
- Never suppresses type errors with
as anyor@ts-ignore - Never deletes failing tests to "pass"
- Never leaves code in a broken state
Platform Compatibility
Squirrel works on: Claude Code, Codex, Cursor, Antigravity, Gemini CLI, GitHub Copilot, Windsurf, OpenCode, Aider (9 total).
Install with:
# Universal installer
npx skills add flyingsquirrel0419/squirrel-skill
Limitations
- Does not replace environment-specific validation or expert review
- CI/CD templates are starting points, not drop-in guarantees
- Parallel sub-agent execution depends on platform support
Related Skills
@brainstorming- For planning before implementation@test-driven-development- For TDD-oriented workflows@systematic-debugging- For methodical problem-solving
04 同仓库更多 Skills
writing-guidelines
Review docs/prose for Writing Guidelines compliance. Use when asked to "review my docs", "check writing style", "audit prose", "review docs voice and tone", or "check this page against the writing handbook".
writing-for-agents
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
wrangler
Cloudflare Workers CLI for deploying, developing, and managing Workers, KV, R2, D1, Vectorize, Hyperdrive, Workers AI, Containers, Queues, Workflows, Pipelines, and Secrets Store. Load before running wrangler commands to ensure correct syntax and best practices. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
workers-best-practices
Reviews and authors Cloudflare Workers code against production best practices. Load when writing new Workers, reviewing Worker code, configuring wrangler.jsonc, or checking for common Workers anti-patterns (streaming, floating promises, global state, secrets, bindings, observability). Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
web-perf
Analyzes web performance using Chrome DevTools MCP. Measures Core Web Vitals (LCP, INP, CLS) and supplementary metrics (FCP, TBT, Speed Index), identifies render-blocking resources, network dependency chains, layout shifts, caching issues, and accessibility gaps. Use when asked to audit, profile, debug, or optimize page load performance, Lighthouse scores, or site speed. Biases towards retrieval from current documentation over pre-trained knowledge.
using-agent-skills
Discovers and invokes agent skills. Use when starting a session, or when you need to decide which skill or workflow applies to the piece of work at hand. This is the meta-skill that governs how all other skills are discovered and invoked.