harness-official-docs-update
Skill0 starsChanged 3 months agoApache-2.0View on GitHub
Use before changing Revharness behavior that depends on current Codex, OpenAI prompt guidance, Claude Code, subagents, skills, hooks, settings, wrappers, or Goal workflows.
In this repository78 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .agents/skills/go-skills-knowledge-pack/AGENTS.mdAGENTS.md
- .agents/skills/rust-skills-knowledge-pack/AGENTS.mdAGENTS.md
- .agents/skills/typescript-skills-knowledge-pack/AGENTS.mdAGENTS.md
- .claude/skills/go-skills-knowledge-pack/AGENTS.mdAGENTS.md
- .claude/skills/rust-skills-knowledge-pack/AGENTS.mdAGENTS.md
- .claude/skills/typescript-skills-knowledge-pack/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .cursor/rules/00-general.mdcCursor rule
- .cursor/rules/revharness-critical.mdcCursor rule
- .cursor/rules/revharness-detailed.mdcCursor rule
- auto-orchestratorSkill
- baseline-protectionSkill
- client-distribution-readinessSkill
- cloudflare-deploy-guardSkill
- codex-app-server-guardSkill
- codex-app-server-product-integrationSkill
- codex-callerSkill
- context-compactorSkill
- cursor-callerSkill
- delegated-development-disciplineSkill
- deploy-command-disciplineSkill
- design-principlesSkill
- development-constitutionSkill
- development-junk-cleanupSkill
- go-skills-knowledge-packSkill
- harness-official-docs-updateSkill
- naming-normalization-guardSkill
- orchestrator-bootstrapSkill
- payload-cms-deploy-guardSkill
- production-function-implementerSkill
- research-handoffSkill
- rev-harness-lifecycleSkill
- review-workflowSkill
- root-cause-first-disciplineSkill
- rust-skills-knowledge-packSkill
- scope-guardSkill
- self-growth-proposal-triageSkill
- shared-client-wrapper-safetySkill
- slice-designerSkill
- staff-code-reviewerSkill
- structured-mentorSkill
- supabase-deploy-guardSkill
- system-plannerSkill
- typescript-skills-knowledge-packSkill
- auto-orchestratorSkill
- baseline-protectionSkill
- client-distribution-readinessSkill
- cloudflare-deploy-guardSkill
- codex-app-server-guardSkill
- codex-app-server-product-integrationSkill
- codex-callerSkill
- context-compactorSkill
- cursor-callerSkill
- delegated-development-disciplineSkill
- deploy-command-disciplineSkill
- design-principlesSkill
- development-constitutionSkill
- development-junk-cleanupSkill
- go-skills-knowledge-packSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 20
Start from docs/official-docs-links.md.
- Distinctive: line 38
Do not let an upstream recommendation silently override local sandbox, approval, wrapper, task-lineage, evidence, or acceptance rules.
- Distinctive: line 45
If Codex app-server Goal transport is added later, it must be explicit opt-in, set the Goal before the run, clear it after the run, and …
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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
Skill: Harness Official Docs Update
This skill owns the upstream-documentation intake step for Revharness behavior changes. Use it before planning or editing harness surfaces whose correctness depends on current OpenAI Codex, OpenAI API prompting, or Claude Code behavior.
Use When
- Updating Codex wrapper, role, model, prompting, Goal, app-server, workflow, or native subagent behavior.
- Updating Claude Code settings, hooks, subagents, skills, permissions, or delegation behavior.
- Simplifying prompts based on GPT-5.5 / Codex / Claude Code guidance.
- Adding or changing a Revharness workflow because an upstream AI coding tool changed.
Official Sources
Start from docs/official-docs-links.md. For Codex / Claude Code surfaces, prefer these official source families:
- OpenAI Codex changelog, workflows, prompting, and app-server docs.
- OpenAI prompt guidance for model/prompt updates.
- Claude Code subagents, settings, hooks, and current subagent reference docs.
Record which pages were consulted in the plan, SOW, or research handoff. If current official docs are unavailable, record the failure and do not claim the update is official-docs-backed.
Local Authority Mapping
Official upstream guidance is input, not Revharness truth. After reading the official source, map the change into the local authority layer:
- runtime wrapper behavior:
scripts/*wrapper*.shand the relevant skill / role docs - durable task contract:
docs/manual/common-task-contract.md - acceptance / LGTM / completion:
docs/manual/verification-truth-matrix.md - role boundaries:
docs/roles/*.md - workflow routing:
.claude/skills/*.md
Do not let an upstream recommendation silently override local sandbox, approval, wrapper, task-lineage, evidence, or acceptance rules.
Goal / Contract Envelope Boundary
- Codex Goal is optional runtime steering or transport only.
- Goal is not a task contract, evidence store, LGTM condition, completion authority, or acceptance truth.
- Non-interactive automation must not inject
/goalslash commands into stdin. - If Codex app-server Goal transport is added later, it must be explicit opt-in, set the Goal before the run, clear it after the run, and fail closed when unavailable.
- Durable Revharness artifacts must continue to carry task lineage, slice id, required checks, evidence destination, completion boundary, and review evidence.
Output
For each update, leave a concise handoff:
- official docs consulted
- upstream behavior or recommendation
- local Revharness authority files affected
- boundaries that were not changed
- deterministic checks required before review
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.

