vkirill/claude-lane-stack/codex-lane
Claude Code and agent skill by vkirill/claude-lane-stack
- community
Checking that the SKILL.md file is still in the repository.
Operate Claude Lane Stack schema-v2 runs from Codex. Use for lane status, start, resume, review, verification, evidence, or acceptance.
Install
npx vskill@latest install vkirill/claude-lane-stack/codex-laneSKILL.md
The same SKILL.md was published earlier as vkirill/claude-lane-stack/codex-lane.
Codex Lane
Use the installed typed wrappers under ~/.agents/bin. Keep the run directory, task file, and project root supplied by the user or existing run state. Do not invent a run, task id, path, model, or receipt.
Dependencies
Check the wrappers before acting when the installation is uncertain:
The wrappers may be invoked by absolute path (~/.agents/bin/...) when PATH does not include ~/.agents/bin. Their lifecycle state lives under the run's .agents/runs/<slug>/ directory.
Run and resume
For an existing schema-v2 run, validate and start the durable controller. Keep the run's configured writer by default; add --provider codex only when Codex is the explicitly selected writer:
run-controller start is idempotent and resumes persisted controller state. Use a fresh bounded watch call after an interrupted session. Read the controller receipt with:
Preserve the run's selected native profile, provider, model, reasoning effort, and service tier unless the operator explicitly asks for a change. Do not silently route a run to Codex or override those values.
For one typed lane instead of the DAG controller (use --provider codex only when explicitly selected):
Status and evidence
Read compact state first, then inspect the bounded journal or a specific log:
Require a trusted complete report.md, matching report_sha256, provider exit 0, and the current attempt before treating a writer as ready for L1. The writer's report is evidence; it is not acceptance.
Review, verify, accept
lane-ctl has no review subcommand. Run the installed read-only review entrypoint when a review is requested:
Run the task's immutable verification and ownership checks:
When the task/run requires review, produce the task-specific receipt before acceptance. Use the immutable pre-task commit recorded by the run as BASE_REF; if it is missing, report the missing baseline. A repository-wide night review does not substitute for this receipt:
Finally apply the typed acceptance gate:
Acceptance must produce artifacts/<task-id>/acceptance.json and a current accepted state. If any receipt is stale or missing, stop and report the exact gate failure instead of editing control-plane artifacts.
Do not run provider commands directly, rewrite report.md, bypass lane-ctl, or claim browser or user-specific verification from a fixture. This plugin does not replace native Codex compaction or add an MCP server.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingplugins/codex-lane/skills/codex-lane/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/vkirill/claude-lane-stack/codex-lane)