jupes/agent-forge-harness / .agents/skills/authoring-agent-skills
authoring-agent-skills
VerifiedCreate or revise Forge skills, including discovery descriptions, workflows, references, and executable helpers.
Authoring Agent Forge skills
Edit .claude/skills/<name>/SKILL.md as the source. .agents/skills/ is a generated
Codex mirror; regenerate it with bun run codex:sync after changes.
Workflow
- Define the job. Identify the requested outcome, inputs, side effects, and nearby skills that could be confused with this one. Reuse the user's context; ask only for decisions that materially change the workflow.
- Write discovery metadata. Use a kebab-case directory and matching
name. Describe the capability and distinct invocation cases indescription. Preserve existing invocation policy; normal Forge skills remain discoverable. - Write the execution path. Give the agent the domain-specific decisions, commands, and evidence needed to finish. Each stage should yield an observable result or explicitly identify what prevents it. Keep shared requirements in the entrypoint; link optional detail at the branch that needs it.
- Package reusable work. Read the authoring guide
when choosing references, helpers, or verification cases. Bundle dependencies
needed by the skill;
repos/contains ignored user clones, not shipped resources. Bun helpers use the repository's JSON envelope. Other runtimes are appropriate for document rendering when their prerequisites are documented. - Verify behavior and discovery. Try an ordinary request, an adjacent request that should select another skill, and a relevant missing-input or failure case. Run changed helpers. Confirm every linked resource exists and every claimed completion has evidence. Structural checks alone do not establish skill quality.
- Refresh consumers. Run
bun run codex:syncandbun run dashboard:commands; inspect the generated diff. Run the applicable repository gates and report their actual results.
For a new folder, the optional scaffold is:
bun run .claude/skills/authoring-agent-skills/scripts/scaffold.ts <skill-name> "<capability and invocation cases>"
Replace every scaffold prompt with working instructions; retain only resources the workflow uses. The finished skill must be usable from a fresh harness clone.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 470k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |