| name | vibeos-autonomous |
| description | Full autonomous-session override for VibeOS in Codex. Use when the user says go autonomous, stop checking in, run on your own, stay in autonomous mode, or wants VibeOS to keep building until a real blocker or decision appears. |
VibeOS Autonomous
Use this skill to turn on a temporary full-autonomous session override. For 24-48 hour work, treat autonomy as a resumable long-run protocol with heartbeat evidence, checkpoints, audit cadence, failure detection, recovery planning, evidence-backed recovery resolution, scheduler guarding, and explicit stop conditions.
Workflow
- Read or create
.vibeos/config.json.
- Preserve the negotiated autonomy level and set:
autonomy.session_override.mode = "autonomous"
autonomy.session_override.active = true
autonomy.long_run.active = true
autonomy.long_run.target_hours = 24
autonomy.long_run.max_hours = 48
autonomy.long_run.heartbeat_interval_minutes = 30
autonomy.long_run.checkpoint_interval_minutes = 60
autonomy.long_run.audit_interval_minutes = 180
- Read or create
.vibeos/session-state.json and mark the session active in autonomous mode with a long_run block.
- Append a mode-change entry to
.vibeos/build-log.md.
- If available, record a heartbeat:
python3 ".vibeos/scripts/autonomy-heartbeat.py" --status running --iteration 0 --summary "autonomous long-run session enabled" --next-action "continue build loop"
python3 ".vibeos/scripts/autonomy-supervisor.py" --project-dir "."
python3 ".vibeos/scripts/autonomy-runner.py" --project-dir "." --json
python3 ".vibeos/scripts/autonomy-loop.py" --project-dir "." --json
python3 ".vibeos/scripts/autonomy-runtime-adapter.py" --project-dir "." --json
python3 ".vibeos/scripts/autonomy-failure-detector.py" --project-dir "." --json
python3 ".vibeos/scripts/autonomy-recovery-planner.py" --project-dir "." --json
python3 ".vibeos/scripts/autonomy-recovery-resolution.py" --project-dir "." --json
python3 ".vibeos/scripts/autonomy-scheduler-guard.py" --project-dir "." --json
python3 ".vibeos/scripts/autonomy-scheduler-profile.py" --project-dir "." --json
python3 ".vibeos/scripts/autonomy-smoke.py" --runtime-provider codex --json
- Explain what changes:
- routine check-ins stop
- 24-48 hour runs write resumable heartbeat, checkpoint, and audit evidence
- gates and audit rules still apply
- the system will still pause for blockers, risk decisions, or plan boundaries
- Continue immediately with
vibeos-build.
Rules
- Keep the override truthful and explicit in state files.
- When autonomous mode ends, clear the override and record the session handoff.
- Use
autonomy-loop.py for scheduler-safe loop ticks when a non-model process is driving the session.
- Respect
.vibeos/autonomy/run-lease.json; a lease_conflict result means another autonomy driver is active.
- Use
autonomy-runtime-adapter.py for dry-run-first Codex/Claude handoff planning; launch only with explicit --execute.
- Use
autonomy-failure-detector.py for repeated handoff loops, runner blocks, runtime failures, lease conflicts, and provider/session limits.
- Use
autonomy-recovery-planner.py for plan-only recovery actions after failure detection.
- Use
autonomy-recovery-resolution.py to record summary and evidence for resolved recovery actions.
- Use
autonomy-scheduler-guard.py to block scheduler-driven ticks while recovery actions lack matching resolution evidence.
- Use
autonomy-scheduler-profile.py only to generate reviewed profile files.
- Use
autonomy-smoke.py before trusting a generated scheduler profile.
- Record a heartbeat at least every 30 minutes, after each WO boundary, and after checkpoint/audit boundaries.
- Classify supervisor resume plans with
autonomy-runner.py; use --execute only for allowlisted local VibeOS scripts.
- Save checkpoints at least every 60 minutes or after each agent/work-package boundary.
- Run checkpoint or session audit evidence at least every 180 minutes or phase boundary.
- Terminal long-run states must record
complete, paused, or blocked through autonomy-heartbeat.py, then pass validate-long-run-autonomy.py --require-closed.