Skip to content

jscraik/Agent-Skills / Infrastructure/references/deferred-skill-context/skill-factory-skillify

skillify

Verified

Capture a completed Codex workflow as a reusable SKILL.md package by analyzing session context plus optional session-collector evidence, interviewing the user with structured prompts, and writing a validated skill artifact. Use when the user asks to skillify or operationalize a repeatable process.

Skillify

Convert a completed workflow into a reusable skill package with explicit invoke syntax.

Read when: intake and template details are needed: skill template Read when: source workflow context comes from prior agent sessions: session collector intake

Philosophy

  • Preserve repeatable behavior, not one-off execution details.
  • Optimize for future operator clarity with minimal ambiguity.
  • Keep generated skills auditable with explicit inputs, outputs, and validation.

When to use

  • Use when a workflow has been repeated enough to justify a reusable skill.
  • Use when the user wants a conversation or run converted into durable SKILL.md guidance.

Required inputs

  • Source workflow context (session transcript, notes, commands used, or ~/.agents/session-collector bundle evidence).
  • Target audience and success criteria for the new skill.
  • Destination path and category for where the skill should live.

Deliverables

  • A complete skill package centered on SKILL.md with clear invoke syntax.
  • Any required companion files referenced by the skill (scripts, templates, references).
  • Validation notes covering contract and structure checks.
  • Structured output includes schema_version: 1 when requested or when automation will consume the handoff.

Procedure

  1. Capture the source workflow and confirm intended reuse scope.
  2. If the workflow source is a prior session or repeated agent run, prefer ~/.agents/session-collector bundle evidence and consume bounded extracted evidence rather than raw transcripts.
  3. Extract stable triggers, required inputs, deliverables, and failure boundaries.
  4. Draft SKILL.md using the template and include explicit procedure/validation/constraints sections.
  5. Add or update supporting references (contract/evals/task profile) needed for governance gates.
  6. Run the relevant structure checks and iterate until clean.

Reference materials:

Constraints

  • Do not codify workflows that are still exploratory or contradictory.
  • Do not inventory all session history inside skillify; use skill-refactor or ~/.agents/session-collector for broad session evidence and consume only the selected workflow extraction.
  • Do not embed repository secrets, credentials, or private user data in generated skill artifacts.
  • Redact sensitive tokens, secret paths, and account identifiers from examples and references.
  • Keep scope aligned to the requested category and destination.
  • Keep the first pass to 2-3 focused surfaces unless the user explicitly asks for a broader skill package.

Validation

  • Verify trigger text maps to realistic user language.
  • Verify required sections and references exist and are internally consistent.
  • Verify eval cases include happy, edge, and failure behavior.
  • Fail fast: if source workflow context is insufficient, stop and report missing inputs.

Anti-patterns

  • Copying raw session transcripts directly into SKILL.md.
  • Generating broad triggers that hijack unrelated tasks.
  • Treating template completion as success without running validation gates.

Failure mode

  • If the workflow is too incomplete or inconsistent, stop and report what is missing.
  • If required destination/category cannot be resolved, pause and request explicit routing.

Gotchas

  • Avoid overfitting to a single run; generalize only repeatable steps.
  • Keep prerequisites explicit so the skill is runnable without hidden assumptions.

Examples

  • "Skillify this repeatable release triage process into a reusable lane under agent-ops."
  • "Convert yesterday's plugin onboarding workflow into a validated skill package."
  • "Use the extracted session notes from the successful PR cleanup run to create a reusable review-fix skill, but keep the raw transcript out of the skill."
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: 468kUpdated Claude CodeCodexVerifiedInstall
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: 321kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 321kUpdated Claude CodeCodexVerifiedInstall
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 availableStars: 321kUpdated Claude CodeCodexVerifiedInstall