← ClaudeAtlas

feature-implementlisted

Execute a decomposed feature via the gsd-core loop with preflight-only host fallback, no stateful cross-vendor replay, autonomy grants, gates.py completion authority, and a fail-closed review/ship tail. --adhoc uses the same walls over gsd-quick.
austinmao/feature-fix-swarm · ★ 3 · AI & Automation · score 72
Install: claude install-skill austinmao/feature-fix-swarm
# feature-implement — Execute a feature through the gsd loop ## Host dispatch contract - Codex: invoke skills as `$skill`; use Codex collaboration roles and GPT-5.6 model tiers. - Claude: invoke skills as `/skill`; use Claude Agent/Skill tools and Claude model aliases. - Examples that name both hosts are routing contracts. Never send one host's command syntax to the other. - A bare `/skill` in this shared source denotes the Claude form; Codex dispatches the same named skill as `$skill`. ## Init gate Before any other step, run the advisory init guard and relay its output verbatim: ```bash bash "$(git rev-parse --show-toplevel)/scripts/gsd/init-guard.sh" || true ``` If it printed `INIT-GUARD:` warnings, offer `/ffs-init` before proceeding in interactive sessions (declining proceeds anyway); headless, spawned, and autonomous runs relay the warnings once and continue. Advisory only — never a block, never an exit-code change. At entry, make one opportunistic, fail-soft `bash scripts/gsd/reconcile.sh` pass; never block on its result. ## When to invoke - After `/feature-spec` or `/spec-decompose` produced a seeded gsd project (`.planning/`) or a legacy `specs/NNN/tasks.md` - "implement NNN", "run tasks for NNN", "execute feature NNN" - **Adhoc (v2.6.0):** a bounded task with NO spec/plan — `--adhoc "<task>"` runs the same walls + gsd loop + finish tail over `/gsd-quick`. `/fix` and `/task-swarm` route here; the delegation machinery lives ONLY in this skill. - Resumes