codex-adapter
Skill1 starsChanged 2 months agoMITView on GitHub
Validate and operate Everything Agentic Engineering with Codex. Use when configuring Codex for this repository, checking AGENTS.md or skill discovery, preparing parallel Codex worktrees, packaging the repository plugin, reviewing Codex hooks, or diagnosing Claude/Codex contract drift.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- wrg32786/aigent-ostheir own version, first seen Oct 13, 2026
- kunchenguid/compact-advisertheir own version, first seen Sep 18, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- kunchenguid/compact-adviser★ 227same name
- wrg32786/aigent-os★ 25same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- accessibility-auditSkill
- agentic-uxSkill
- benchmarkSkill
- capability-routerSkill
- context-handoffSkill
- create-prdSkill
- creative-direction-sprintSkill
- decompose-prdSkill
- design-criticSkill
- design-engineering-qualitySkill
- design-intakeSkill
- design-opsSkill
- design-system-auditSkill
- design-systemSkill
- design-tokensSkill
- discoverSkill
- harness-evolutionSkill
- interaction-designSkill
- loop-engineeringSkill
- parallel-planSkill
- performance-uxSkill
- polishSkill
- product-design-routerSkill
- project-onboardingSkill
- research-ledgerSkill
- responsive-auditSkill
- security-gateSkill
- self-improvement-loopSkill
- specialist-routerSkill
- strategizeSkill
- token-auditSkill
- user-needsSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
16 rules. About 0 are common, 7 are distinctive, and 9 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
9
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Very specific: line 10
Read AGENTS.md, .agentic/project.json, the active task, and docs/60-tooling/CODEX.md.
- Very specific: line 12
Run ./scripts/codex-doctor.sh before changing Codex-specific files.
- Very specific: line 13
Treat AGENTS.md as the cross-runtime constitution.
- Distinctive: line 16
Do not add model selection, credentials, provider routing, network access, MCP servers, sandbox expansion, or approval bypasses without an …
- Distinctive: line 22
Use the local checkout for orchestration or one foreground task.
- Very specific: line 23
Use .codex/agents/ for bounded read-only product, architecture, research, design, security, QA, and integration analysis.
- Very specific: line 25
Prefer Codex's built-in explorer for general codebase mapping rather than duplicating that role locally.
- Distinctive: line 27
Use one branch and worktree per independent write-heavy task.
- Distinctive: line 31
Keep the final evaluator separate from the builder.
- Distinctive: line 32
Do not add a write-capable project subagent merely to simulate worktree isolation inside one checkout.
- Very specific: line 35
Use ./scripts/task-plan.sh <TASK-ID> before workspace creation and follow docs/70-collaboration/PARALLEL_TERMINALS.md for multi-terminal …
- Very specific: line 40
Keep .agents/skills and the plugin skills path pointed at the canonical .claude/skills tree.
- Very specific: line 42
Keep .codex/hooks.json wired to the reviewed scripts under .claude/hooks/.
- Very specific: line 44
Keep every committed .codex/agents/*.toml role read-only and free of model, provider, MCP, network, credential, or approval-policy choices.
- Distinctive: line 48
Update compatibility documentation when a runtime floor or limitation changes.
- Distinctive: line 51
Read references/contract.md when changing adapter files, packaging, or acceptance criteria.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Codex adapter
Establish the runtime contract
- Read
AGENTS.md,.agentic/project.json, the active task, anddocs/60-tooling/CODEX.md. - Run
./scripts/codex-doctor.shbefore changing Codex-specific files. - Treat
AGENTS.mdas the cross-runtime constitution. Treat.codex/as the Codex adapter and.claude/as the canonical local skill and safety-script source. - Do not add model selection, credentials, provider routing, network access, MCP servers, sandbox expansion, or approval bypasses without an explicit, separately reviewed requirement.
Route work safely
- Use the local checkout for orchestration or one foreground task.
- Use
.codex/agents/for bounded read-only product, architecture, research, design, security, QA, and integration analysis. - Prefer Codex's built-in
explorerfor general codebase mapping rather than duplicating that role locally. - Use one branch and worktree per independent write-heavy task.
- Assign exclusive file or tightly coupled module ownership before parallel work starts.
- Merge shared contracts before dependent implementations.
- Keep the final evaluator separate from the builder.
- Do not add a write-capable project subagent merely to simulate worktree isolation inside one checkout.
Use ./scripts/task-plan.sh <TASK-ID> before workspace creation and follow
docs/70-collaboration/PARALLEL_TERMINALS.md for multi-terminal work.
Maintain the adapter
- Keep
.agents/skillsand the pluginskillspath pointed at the canonical.claude/skillstree; do not create hand-maintained copies. - Keep
.codex/hooks.jsonwired to the reviewed scripts under.claude/hooks/. - Keep every committed
.codex/agents/*.tomlrole read-only and free of model, provider, MCP, network, credential, or approval-policy choices. - Validate changes with
./scripts/codex-doctor.sh, the skill validator, the plugin validator, and./scripts/verify.sh full. - Update compatibility documentation when a runtime floor or limitation changes.
Read references/contract.md when changing adapter files, packaging, or acceptance criteria.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.