Workflow skill
Dev Lifecycle Skill
Orchestrate requirements, design, planning, implementation, testing, and review.
On this page
Install command
AI DevKit installs its built-in skills as one curated set. Run this once to install dev-lifecycle with the rest of the built-in workflow skills.
npx ai-devkit@latest skill add --built-inWhat it does
Use dev-lifecycle when you want an AI coding agent to follow a structured software delivery process instead of improvising in chat. It routes work to focused phase skills and keeps higher-risk engineering work moving through explicit checkpoints.
When to use it
- Run a complete feature lifecycle with one repeatable workflow
- Choose the next delivery phase across requirements, design, planning, implementation, testing, and review
- Coordinate phase-specific AI DevKit skills without manually remembering every step
Supported agents
- Claude Code
- Codex
- Cursor
- opencode
- Gemini CLI
The SKILL.md file
This is what your agent reads when the skill runs. View it on GitHub (opens in a new tab)
Coordinate the phase-specific AI DevKit skills instead of running phase details directly.
Required phase skills:
dev-worktreefor feature workspace setup and resume.dev-requirementsfor phases 1-2: new requirement and requirements review.dev-designfor phase 3: design review.dev-planningfor phases 4 and 6: initial task planning and updates after implementation tasks.dev-implementationfor phases 5 and 7: execute plan and check implementation.dev-testingfor phase 8: write tests and verify coverage.dev-reviewfor phase 9: final code review.
Supporting skills:
memoryfor reusable project knowledge during clarification.- Load
tddwith your skill tool before writing code in phase 5. - Load
verifywith your skill tool before any completion claim in phases 5, 7, 8, and 9, and immediately before every commit. - Load
dev-smokewith your skill tool after Write Tests and before Code Review when the change touches a user-facing entry point (CLI command, API route or RPC handler, UI screen, daemon or wire protocol). Skipping it requires a stated reason. - Load
dev-simplifywith your skill tool in phase 7 for an analysis-only pass. Apply its changes only if it finds something worth fixing and the user or plan allows. - Load
dev-commitwith your skill tool before committing. dev-retroafter Code Review or PR: offer it, and run only if the user agrees.taskfor optional progress tracing when the task command is usable.
Following a summary of a skill is not a substitute for loading it.
Startup Validation#
At the beginning of every dev-lifecycle run:
- Run
npx ai-devkit@latest skill listandnpx ai-devkit@latest skill list --globalto inspect currently installed project and global skills. - Confirm the required phase skills and supporting skills are covered by the union of both lists.
- Only if a required skill is missing from both, run
npx ai-devkit@latest skill add --built-into install all AI DevKit built-in skills. Then rerun both lists. - If installation fails or a required skill is still missing from both, stop and report the missing skill names and command output summary. Do not run a phase without its skill.
- Run
npx ai-devkit@latest lintto verify the configured AI docs structure. - If working on a specific feature, run
npx ai-devkit@latest lint --feature <name>. - If lint fails because project docs are not initialized, run
npx ai-devkit@latest init -a -e claude --built-in --yes, then rerun lint. - Probe optional task tracing availability:
- With a feature:
npx ai-devkit@latest task list --name <feature-name> --json - Without a feature:
npx ai-devkit@latest task list --json - Treat task tracing as available only if the read probe exits 0. If it fails, record task tracing as unavailable with the failed command and reason, then continue without task logging.
- Never block lifecycle work only because the task command is missing or unusable.
- With a feature:
- When working on a specific feature and task tracing is available:
- Load and follow
taskbefore executing a phase. - Initialize or show the task named after the feature, mark active work, and emit phase/progress/next/blocker/evidence events per
task. - Sequence task mutations; do not batch or parallelize mutations for the same feature.
- Load and follow
Plan Before Execution#
Before executing any phase:
- Identify the target feature, current docs state, branch/worktree context, and likely next phase.
- Propose a concise plan that names the phase skill to use, the docs/files to read, commands to run, expected edits, task tracing status and planned task events if tracing is available, and verification evidence.
- Wait for user approval before executing the plan unless the user already gave explicit approval for that exact phase execution.
- After approval, load and follow only the selected phase skill plus any explicitly required supporting skills. If tracing is available, the
taskskill is explicitly required.
Phase Routing#
| Phase | Route to | When |
|---|---|---|
| Setup. Workspace | dev-worktree | Starting or resuming feature work |
| 1. New Requirement | dev-requirements | User wants to add a feature or start /new-requirement |
| 2. Review Requirements | dev-requirements | Requirements doc needs validation |
| 3. Review Design | dev-design | Design doc needs validation against requirements |
| 4. Create Initial Plan | dev-planning | Requirements, design, and testing docs are ready for task breakdown |
| 5. Execute Plan | dev-implementation | Ready to implement tasks from planning doc |
| 6. Update Planning | dev-planning | Auto-trigger after completing any implementation task |
| 7. Check Implementation | dev-implementation | Verify code matches design and docs |
| 8. Write Tests | dev-testing | Add or verify test coverage |
| 9. Code Review | dev-review | Final pre-push review |
Sequential flow: setup -> 1 -> 2 -> 3 -> 4 -> 5 -> 6 after each completed task -> 7 -> 8 -> 9.
Required Skill Loads#
| Phase | Load with your skill tool |
|---|---|
| 5. Execute Plan | dev-implementation, tdd |
| 7. Check Implementation | dev-implementation, verify, dev-simplify (analysis) |
| 8. Write Tests | dev-testing, verify |
| Before 9 | dev-smoke when the change touches a user-facing entry point |
| 9. Code Review | dev-review, verify |
| Commit | dev-commit, verify |
End every phase summary with Skills loaded: <skills> and Skipped: <skill> (<reason>). When task tracing is available, record that line with task progress.
Resuming Work#
If the user wants to continue work on an existing feature:
- Use
dev-worktreeto identify and confirm the target branch/worktree. - Run
npx ai-devkit@latest lint --feature <feature-name>in the active context. - Run the phase detector from the installed
dev-lifecycleskill directory:- Resolve
<skill-dir>as the directory containing thisSKILL.md. - Run
<skill-dir>/scripts/check-status.sh <feature-name>. - Use the suggested phase when proposing the execution plan.
- Resolve
Backward Transitions#
Not every phase moves forward. When a phase reveals problems, route back:
- Requirements review finds fundamental gaps: return to
dev-requirementsPhase 1. - Design review finds requirements gaps: return to
dev-requirementsPhase 2. - Design review finds design flaws: stay in
dev-designand revise design. - Implementation check finds major deviations: return to
dev-designif design is wrong, ordev-implementationif code is wrong. - Testing reveals design flaws: return to
dev-design. - Review finds blocking issues: return to
dev-implementationordev-testing. - Smoke check finds a regression: return to
dev-implementation. - User requests changes after review: return to
dev-implementation(or the earlier phase the change needs).
When task tracing is available, record every backward transition with task phase, and with task feedback when it comes from user or review feedback (see task Rework).
Rules#
- Use
npx ai-devkit@latest lintandnpx ai-devkit@latest lint --feature <name>to discover and validate the configured docs directory. Do not assumedocs/ai; it is only the default. - Read existing configured AI docs before changes. Keep diffs minimal.
- Keep feature names aligned with branch/worktree
feature-<name>. - New feature docs come from
npx ai-devkit@latest docs init-feature <name>. Use the paths returned by the command as authoritative. - Existing feature docs are the paths reported or validated by
npx ai-devkit@latest lint --feature <name>. If you must infer manually, first resolve the configured docs directory from.ai-devkit.jsonpaths.docs, falling back todocs/ai. - After each phase, summarize output and suggest the next phase.
- Pre-approval removes approval waits; it never skips phases or required skill loads.
- Do not claim completion without fresh verification evidence.
- When task tracing is available, follow
task: create once, assign actor when known, mark active/blocked, set phase, record progress/next/evidence, and close only after final verification/review. If tracing is unavailable, include failed probe commands in the phase summary without blocking the lifecycle.