Install
$ agentstack add skill-davidvictor-adlc-skills-adlc-implement ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
ADLC Implement
Use this to execute an existing plan or fix plan.
Process
- Resolve effective ADLC paths from
.adlc/config.yaml; useadlc resolve-configwhen the CLI is available. - Locate the plan. Support explicit path, task number,
status, or current branch. - Read any interview spec cited by the plan or workstream before editing.
- If executing a workstream step, read
WORKSTREAM.md,kanban.md, the relevant milestone card, and the selected step card before editing. - For goal-managed workstreams, check the active Codex goal at the start and preserve the goal objective through closeout.
- Read configured skill-context before fallback patch history.
- Use
implement-coordinatorwhen native Codex agents are available. - Keep tightly coupled work local in the coordinator.
- Delegate independent bounded edits to
implement-workerwith disjoint write scopes. - Run read-only sidecars after meaningful code changes: review, security, rules, docs, best-practices, and commit-preparer as relevant.
- Integrate worker results, rerun verification, and update plan, milestone, or workstream step status.
Rules
- Do not exceed the selected task scope.
- Do not cross interview non-goals or decision boundaries without recording the blocker.
- Do not advance a workstream step past
commituntil the committed slice is explicit. - Do not mark a Codex goal complete until the whole workstream objective is achieved.
- Do not let sidecars edit files.
- Do not call work complete without verification and a final integration state.
Output
Report completed tasks, changed files, verification, sidecar findings, residual risk, workstream or goal status when relevant, and the next command: adlc-verify, adlc-review, adlc-commit, or adlc-fix.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: davidvictor
- Source: davidvictor/adlc-skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.