Menu
Route software work through the Agentic Development Lifecycle in Pi. Use when the user asks to apply ADLC, operate in ADLC, triage work, or choose the next ADLC gate.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
| name | adlc |
| description | Route software work through the Agentic Development Lifecycle in Pi. Use when the user asks to apply ADLC, operate in ADLC, triage work, or choose the next ADLC gate. |
Classify the work before acting:
Use deterministic ADLC CLIs for pass/fail. Skills may recommend commands; they do not declare gates complete unless the relevant CLI and .adlc/manifest.jsonl evidence pass.
Start with:
adlc preflight --json
For strict phase assertions, use:
adlc run <p1|p2|p3|p4|p5|p6|p7> --dir .adlc --json
P3, P4, P5, and P6 assertions are ticket-scoped; include --ticket <ticket-id> for those phases.
Reference adversarial-review at P1 (design review of the ticket/spec) and P3
(attack the declared rail set for adequacy — is every invariant covered and
unbypassable) — recommended practice today via --prompt-only or diff review, since it
only reviews a git diff/branch and first-class artifact input (--input) is a deferred
follow-on — and at P5 (built code; ≥2 distinct providers on the risk gate), where it
runs directly. It is a cross-model, fresh-context ship/no-ship review that loops
review→fix→re-review until clean (exit 0 = SHIP). Flags: --verify (refute stale
findings), --loop (autonomous fix loop over working-tree code changes only, needs a
write sandbox), --providers (multi-provider quorum). Installed separately — invoke via
npx adversarial-review if not on PATH. See ADR-0008 (adversarial-review coverage map)
in the ADLC repo.
Refute-charter review discipline for prosecuting a diff against its ticket. Use when asked to act as a prosecutor, reviewer, or to find reasons a change must not merge.
Run ADLC P7 distillation and maintenance workflows in Codex, including lesson-foundry, rejection-mining, skill-mining, scheduled maintenance, skill-rot, model-ratchet, review-calibration, and gate-fuzzing.
Run the multi-lens P5 adversarial prosecution fan-out and record ADLC P5 review evidence and P6 acceptance packet workflows in Codex. Fans out five independent read-only lenses, dedupes findings, verifies each independently, and loops until dry before recording evidence.
Run ADLC P7 distillation in the GitHub Copilot CLI — lesson-foundry, rejection-mining, and review-calibration — turning repeated review findings and PR rejections into permanent, deterministic defenses. Decay-driven maintenance lives in the adlc-maintain skill.
Run the decay-driven ADLC maintenance checks — stale skills, hot files to re-prosecute, stale tickets, and gate calibration (C10/C12).
Run the multi-lens P5 adversarial prosecution fan-out and record ADLC P5 review evidence and P6 acceptance packet workflows in the GitHub Copilot CLI. Fans out five independent read-only lenses, dedupes findings, verifies each independently, and loops until dry before recording evidence.