gaia-research/gaia-skill-tree / .agents/skills/gaia-steward-land
gaia-steward-land
VerifiedLand a Gaia Steward Class A repair (agent-skill mirror or bundled-schema mirror) as a deterministic commit on a steward/a/* branch, open a draft PR to the integration branch, and record the outcome. Use when a full Steward run emits a forge.open_pr action, for "land the class A repair", or "open the steward PR". Never targets main and never merges.
gaia-steward-land
Responsibility
Turn one verified Class A repair into a reviewable branch and draft PR on the integration branch. The runner builds the commit; this skill only opens the PR and records it.
Only forge.open_pr is landed here. An issue closure (forge.close_issue) is not a
landing and is never performed by this skill; hand it to gaia-steward-issue-hygiene. A routine pull-request merge or
closure (forge.merge_pr, forge.close_pr) is not a landing either; hand it to gaia-steward-pr-maintenance.
Input
A pending forge.open_pr action from gaia steward full run. Its params give
the branch (steward/a/<kind>-<fingerprint>), the base (the integration
branch) and a PR body file under .gaia/steward/full/<run>/.
Output
A draft PR from params.branch to params.base, recorded with:
gaia steward full record RUN STEP --input pr.json
where pr.json is {"pullRequest": {"number": 123, "url": "..."}, "merged": false}.
Authority ceiling
- Only the two Class A executors in
POLICY.yamlproduce landings. - The commit touches only the executor's writable mirror path, proven by a parity check of the committed tree.
- Base is the configured integration branch; the runner refuses
main,masterand the remote default. No force push, no merge.
Stop conditions
- The run reports unknown coverage: no landing this run.
params.baseis not adev/*integration branch: refuse and report.- A PR for the branch already exists: record that PR, do not open a second.
Idempotency
The branch name binds the debt and the canonical digest, so the same drift
lands once. Re-running reports already-exists.
Failure states
blocked (dirty plan, parity mismatch, protected base, push refused) or
already-exists. Neither is retried automatically.
Validation
python -m pytest tests/steward/test_landing.py tests/steward/test_repairs.py -q
Tests
tests/steward/test_landing.py, tests/steward/test_fullrun_scenarios.py,
tests/steward/test_steward_skills.py.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |