| name | talos-ops-sweep |
| description | Use when a Talos change, project pass, or failing build needs one coordinated sweep across CI/build triage, UI surface parity, and dirty-worktree or submodule hygiene before implementation starts. Run the three automation passes, merge the findings, and route the next work to the right owning surface. |
Talos Ops Sweep
Use this skill when the recurring Talos workflow is "inspect first, classify
the real surface, then route the fix".
Load:
../planning/program-anchor-index.md../talos-ci-triage/SKILL.md../talos-ui-surface-parity/SKILL.md../talos-submodule-hygiene/SKILL.md
Workflow:
- Start with the dirty surface. If the repo or submodules are dirty, run the
hygiene pass first so root-level noise does not hide the real scope.
- If a build or CI failure exists, reduce it to the first actionable signal
with the CI triage pass before picking any repro commands.
- If dashboard, API Workbench, or TUI work is in scope, inventory those
surfaces and classify parity gaps before proposing implementation.
- Merge the three outputs into one routing packet: owned surface, failure or
parity class, cleanup actions, and the smallest next verification step.
- Only then choose the implementation skill or specialist agent for the next
lane.
Helper:
python3 .agent/skills/talos-ops-sweep/scripts/run_ops_sweep.py --repo-path . --submodules --format markdown- Add
--ci-log path/to/log.txt when CI or local failure logs are available.
Guardrails:
- Do not jump straight to implementation when the dirty surface, failing
surface, or UI ownership is still ambiguous.
- Do not let one noisy CI log line override submodule or worktree evidence.
- Do not mix confirmed findings with speculative roadmap items.
Done checklist:
- Dirty surface classified.
- CI signal reduced or explicitly marked absent.
- UI surface inventory included when relevant.
- Next owner skill or specialist agent is named.
