Use when the user explicitly asks for $threads, Codex-native subagents, 开几个子 agent, or a GitHub issue/PR queue needing parallel lanes, worktrees, review/merge gates, and closure audit.
SkillScore76Maintained
Automatic
◆AI Engineer pack
Last verified 9/3/2026
What it changes
Passed 15 of 20 test cases with the skill, and 4 of 20 without it.
Without skill20%
With skill76%
Checks it passed that the model alone missed
Classifies the remote truth as level B.
Forbids executing PR merges under truth level B.
Lists shared configuration files in forbidden_files.
and 19 more in the full report
15 fixed4 passed either way5 still failing1 broke
Pass-rate lift
+56 pp
Tokens
+351%
Agent turns
0%
Response time
-29%
Model gemini-3.6-flashMeasured 9/3/2026v219 of 25 cases passed with skill · 20 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 15 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 15
Prompt
A worker subagent fixed lint errors in PR #80, but the reviewer found a secondary style issue in the same class of findings. The fix was re-applied, and the reviewer flagged it a third time. What circuit breaker rule applies?
What was checked, without and with the skill
Checked
Without skill
With skill
Stops execution with failure code REVIEW_LOOP after repeated fix and review cycles.
Missed
Passed
case-22 · gemini-3.6-flash · 9/3/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md605 lines7 filesincludes scripts
Threads
Use this skill to turn a broad request into controlled Codex-native subthreads with explicit lanes, file ownership, review gates, and verifiable closure.
Native Codex threads are short-lived parallel work lines inside the Codex workflow. They are not the same as OMX/tmux workers. If native subagent tools are not visible, discover them with tool search. If no native subagent capability is available, produce the thread prompt pack and execution plan instead of pretending threads were launched.
Quick Path
Classify the request: single_agent, plan_only, execute_direct, review_only, research_spec, or clarify_first.
If the user explicitly asked for threads, record intent_contract and thread_dispatch_gate, write a run_phase: preflight record, and pass append_run_log.py --validate-only before spawning.
For GitHub queues, fetch remote state in the coordinator lane, then write queue_gate, queue_ledger, issue_to_pr_map, and the validated queue_bounds.
Write a lane map with file ownership, verification owner, stop conditions, context budget, and output firewall.
Route large command output to artifacts before dispatch; the parent reads only summaries, short tails, targeted greps, and evidence paths.
Dispatch only bounded native lanes with disjoint writable scopes or read-only roles.
Collect, wait, and close spawned agents; do not count the coordinator as a spawned thread.
Run fresh verification tied to the current head or artifact.
End with a compact final report and threads_run_log; write durable JSONL for queue, multi-lane, push, comment, or merge-capable runs.
Do Not Use For
Do not use this skill for generic uses of "thread" unless the user explicitly means Codex workflow orchestration:
operating-system threads, language concurrency, or async programming models
chat, email, forum, Slack, GitHub discussion, or comment threads
OpenAI Assistants API threads or other product APIs
ordinary single-agent repo work where there are no independent lanes
Decision
Choose one mode:
single_agent: handle a small, well-scoped task locally with the same evidence gates.
plan_only: map issues, PRs, risks, and parallelization without edits.
execute_direct: run one or more bounded implementation lanes after planning.
review_only: launch independent reviewers for PRs, diffs, or risky code.
research_spec: split exploration by angle, then synthesize docs/spec/issues.
clarify_first: ask only when repo, target queue, permission, or done-when is missing.
Use single_agent only for one-file fixes, simple questions, or tasks where the next step depends on one immediate result. If the user explicitly asks for threads, subagents, or a GitHub issue/PR queue and native subagents are available, do not silently choose single_agent; record a concrete no_spawn_reason before implementation work begins.