fixflow
Fixflow orchestrates end-to-end coding delivery with full upfront planning, sequential step execution, and continuous validation through testing. Use this skill when you need strict workflow control with test coverage before handoff, configurable commit policies (default per-step, or final-only/milestone on request), and optional behavior-driven development for ambiguous requirements or when explicitly requested by users.
git clone --depth 1 https://github.com/majiayu000/spellbook /tmp/fixflow && cp -r /tmp/fixflow/skills/fixflow ~/.claude/skills/fixflowSKILL.md
# Fixflow ## Overview Use this skill to deliver end-to-end engineering work in one run: - Plan fully. - Execute in strict sequence. - Validate continuously. - Use explicit commit policy (`per_step` default; `final_only`/`milestone` when requested). - Add BDD behavior specs when requested or when requirements are ambiguous. ## Trigger Cues Trigger this skill when the user asks for one or more of: - Full plan + sequential execution. - No backward compatibility. - Test everything before handoff. - Commit after completion or commit before each next step. - Behavior-driven delivery (BDD) or acceptance scenarios. ## Workflow ### 1. Define Ready Criteria (DoR) > See shared delivery base: [`references/delivery-base.md`](./references/delivery-base.md) ### 2. Build Complete Plan Before Editing > See shared delivery base: [`references/delivery-base.md`](./references/delivery-base.md) ### 3. Add BDD/TDD Layer When Needed Use BDD if user requests it, or if requirements are unclear. #### 3.0 TDD Mode (RED -> GREEN -> IMPROVE) When user requests TDD or test-driven delivery: 1. **RED** — Write a failing test first - Extract testable behavior from requirements - Write minimal test case asserting expected behavior - Run test, confirm it fails (red) - If test passes unexpectedly -> requirement already met or test is wrong 2. **GREEN** — Write minimal implementation - Write just enough code to make the test pass - No extra improvements, no "while I'm here" changes - Run test, confirm it passes (green) 3. **IMPROVE** — Refactor under green tests - Eliminate duplication, improve naming, simplify logic - Run tests after each refactor to confirm no regression - Refactoring must not change external behavior Coverage target: 80% line coverage for new code, 100% for critical paths. Trigger cues for TDD mode: - User says "TDD", "test-driven", "test first" - User says "write tests before code" - Requirements are well-defined with clear inputs/outputs For BDD Lite, Scenario Quality Checklist, Scenario Outline, and Test Layer mapping, see the shared reference: > [`references/bdd-guide.md`](./references/bdd-guide.md) ### 4. Execute Step by Step > See shared delivery base: [`references/delivery-base.md`](./references/delivery-base.md) For `per_step` (default): - Stage only files for current step. - Run step-level checks. - Commit immediately after step checks pass. - Record step -> commit hash mapping. ### 5. Apply No-Backward-Compatibility Mode (When Requested) > See shared delivery base: [`references/delivery-base.md`](./references/delivery-base.md) ### 6. Validate with Test Matrix (Verification Loop) > See shared delivery base: [`references/delivery-base.md`](./references/delivery-base.md) ### 7. Commit and Handoff > See shared delivery base: [`references/delivery-base.md`](./references/delivery-base.md) ## Guardrails - Do not stop at planning when implementation is expected. - Do not leave partially completed plan steps. - Do not defer required testing when it can be run now. - Do not default to `final_only`; use `per_step` unless user explicitly asks otherwise. - Do not move to the next plan step before committing when `commit_policy = per_step`. - Do not claim compatibility if user explicitly requested no compatibility work. - Do not include unrelated pre-existing dirty files in commits. - Do not hand off without concrete validation evidence.
Audit AND optimize a CLAUDE.md / AGENTS.md instruction file — score it against the five high-leverage patterns, flag anti-patterns, then apply approved fixes in place. Use when the user says 优化 CLAUDE.md / 优化 AGENTS.md / optimize my agent doc / 帮我改 claudemd, or after an audit when they want the fixes applied (not just reported).
Generate or update repository-specific AGENTS.md instruction files from real repo evidence. Use when asked to create, design, scaffold, split, or improve root or scoped AGENTS.md files for Codex/Claude/agent workflows, especially when a repo needs directory-specific rules, validation commands, generated-file boundaries, or a short agent onboarding router.
REST/GraphQL/gRPC API design best practices. Use when designing APIs, defining contracts, handling versioning. Covers OpenAPI 3.2, GraphQL Federation, gRPC streaming.
Mobile app UI design expert for iOS and Android. Use when designing app interfaces, creating design systems, ensuring accessibility, or following platform guidelines. Covers Material Design 3, Human Interface Guidelines, color theory, typography, and 2025 trends.
End-to-end app feature inventory and user-story testing workflow with a canonical tracker. Use when the user asks to audit every feature, derive expected behavior from code, test user journeys, or explicitly fix and retest documented UX or logistical defects.
Design architecture foundations before implementation. Use when asked to design or refactor architecture, choose Rust/Go crate, package, module, runtime, workflow, or service boundaries, compare mature project architecture, prevent stacked one-off PRs, audit migration debt in existing repos, or produce a spec before coding.
Evidence-driven architecture research for understanding real systems and making technical decisions. Use when doing architecture landscape studies, source-backed system archaeology, build-vs-buy or adopt/adapt/build decisions, open-source and commercial comparisons, revisiting an earlier architecture choice, or handling requests such as 架构调研, 架构选型, 竞品架构, 技术尽调, 同类方案, 开源替代, how is X built, and what should we learn from X. Do not use for small mechanical changes, market-only discovery, or detailed design after the technology direction is already fixed.
Ask Grok or Gemini via opencli (browser-session based, no API cost). Use when the user says "问 grok", "问 gemini", "ask grok", "ask gemini", "用 grok 查", "grok 怎么看", "让 gemini 分析", or wants a second opinion from Grok/Gemini without paying API tokens. Routes to opencli which drives the already-logged-in Chrome session.