ulw-loop

Use this skill when the user asks for ulw-loop, ulw, durable goal execution, evidence-led work, manual QA, or checkpointed long-running delivery.

This skill is intentionally compact. The full workflow lives in references/full-workflow.md. Read only the sections needed for the current phase, then execute them exactly.

Required First Steps

1. Open references/full-workflow.md.
2. Read through Bootstrap (including its tier triage), Execution Loop, the Manual-QA channels table, and the Stop Rules before running any ULW command or recording evidence.
3. Open references/define-goal.md and register the run's goal by it. Goal creation is NEVER skipped: shape the objective and every success criterion by that reference before any implementation.
4. If the task has code edits, tests, QA, or commit work, follow the full workflow's delegation and evidence rules. Tests alone never prove done.

Non-Negotiables

• Use the ulw-loop CLI state under .omo/ulw-loop; do not hand-edit goal state.
• Register goals up front, shaped by references/define-goal.md (omo-agent-toolkit ulw-loop create-goals, then create_goal from the printed handoff), and mirror every atomic step into the live update_plan checklist: one ultra-granular step per action, exactly one in_progress, transitions marked the instant they happen. Use the ultrawork directive’s Task resources and todo rules; read those sections of the bundled ultrawork skill if absent from context.
• After any compaction or context loss, re-read brief + goals + ledger FIRST plus omo-agent-toolkit ulw-loop status --json, then resume; never re-plan from scratch.
• Every ulw-loop command needs the session scope: pass --session-id <id> (the printed handoff and resume directive carry it; CODEX_THREAD_ID in the environment also resolves it). The CLI refuses unscoped state (ULW_LOOP_SESSION_SCOPE_REQUIRED) instead of touching the shared .omo/ulw-loop root.
• If omo-agent-toolkit ulw-loop create-goals says this session's aggregate is already complete, start unrelated new work with a fresh --session-id <new-id> (passed on every later call) instead of steering or forcing the completed state. Use --force only to intentionally overwrite completed evidence.
• Every success criterion needs observable evidence from a real surface: a channel (terminal/TUI via the xterm.js web terminal, HTTP, browser, computer-use) or, for CLI- or data-shaped criteria, an auxiliary surface (CLI stdout, DB diff, parsed config dump).
• Evidence is bound to the tree it was captured at (git rev-parse --short "HEAD^{tree}"); it goes stale only when tracked content changes — a rebase or amend that keeps the tree identical keeps it valid. When the tree differs, re-run at the current HEAD and re-record, never relabel or regenerate. Record only after cleanup receipts exist.
• Delegate code edits, test writes, fixes, and QA execution to right-sized Codex subagents when the workflow requires it.
• Every spawn_agent message starts with TASK:, then names DELIVERABLE, SCOPE, and VERIFY; put role and specialty instructions inside message; use fork_turns: "none" (v1: fork_context: false) unless full history is truly required.
• Plan and reviewer agents may run for a long time; spawn them in the background and keep doing independent root work. Between wait_agent calls, back off — double the timeout up to ~5 minutes — instead of spinning short cycles.
• For work likely to exceed one wait cycle, require the child to send WORKING: <task> - <current phase> before long reading, testing, or review passes, and BLOCKED: <reason> only when it cannot progress.
• Track spawned agent names locally. Use wait_agent for mailbox signals, not proof of completion. A timeout only means no new mailbox update arrived. Treat a running child as alive.
• While children run, surface the active subagent count, agent names, and latest WORKING: phase.
• Fallback only when the child is completed without the deliverable, ack-only after followup_task, explicitly BLOCKED:, or no longer running. Then record inconclusive and respawn a smaller fork_turns: "none" task with the missing deliverable.

…