Skip to content

JPeetz/agent-skills / skills/skill-development/skill-miner

skill-miner

Verified

Use this skill when mining agent history for repeated, skill-worthy workflows that deserve to become reusable skills.

FileKindSize
CHANGELOG.mdmarkdown404 B
LICENSEother1.1 KB
SKILL.mdskill md3.9 KB
evals/evals.jsonconfig3.1 KB
references/discovery-rubric.mdmarkdown5.8 KB
scripts/scan_sessions.pyscript7.7 KB

evals/evals.json

View on GitHub
{
  "skill": "skill-miner",
  "version": "1.0.0",
  "cases": [
    {
      "id": "e1",
      "description": "Should trigger: mining session history for workflows",
      "input": "Scan my last month of agent sessions and find patterns that should become skills.",
      "shouldTrigger": true,
      "expect": "Locates session archives, clusters repeated workflows, produces candidate list",
      "assertions": [
        "Searches for session/archive history",
        "Clusters repeated work by intent and tools used",
        "Produces a shortlist with recurrence counts"
      ]
    },
    {
      "id": "e2",
      "description": "Should trigger: identifying a repeated manual workflow",
      "input": "I keep asking you to format dates the same way every time. Make it a skill.",
      "shouldTrigger": true,
      "expect": "Recognizes this as a mining opportunity, drafts skill outline",
      "assertions": [
        "Checks for recurrence evidence",
        "Produces a named skill candidate with trigger description"
      ]
    },
    {
      "id": "e3",
      "description": "Should trigger: team backlog of candidates from actual work",
      "input": "We want a backlog of candidate skills based on our team's real agent usage this quarter, not brainstorming.",
      "shouldTrigger": true,
      "expect": "Runs the scan-sessions.py first pass and scores candidates by recurrence, friction, risk, portability, value",
      "assertions": [
        "Uses scripts/scan_sessions.py for a first-pass cluster report",
        "Scores candidates on the 1-5 recurrence/friction/risk scale",
        "Recommends personal vs public direction per candidate"
      ]
    },
    {
      "id": "e4",
      "description": "Should trigger: recurring procedure in project notes and scripts",
      "input": "Our repo scripts and runbooks keep repeating the same release workflow. Is that skill-worthy?",
      "shouldTrigger": true,
      "expect": "Treats repeated scripts/notes as evidence, checks 3+ occurrences before drafting",
      "assertions": [
        "Checks for 3+ repeat instances before drafting",
        "Filters one-off tasks and project-specific AGENTS.md content",
        "Cross-checks against skill-lifecycle-foundry for gaps"
      ]
    },
    {
      "id": "n1",
      "description": "Not trigger: single one-off task",
      "input": "Help me write a shell script to rename CSV headers.",
      "shouldTrigger": false,
      "expect": "Does not enter the mining workflow",
      "assertions": [
        "Does not invoke the full mining procedure",
        "Treats as a one-off coding task"
      ]
    },
    {
      "id": "n2",
      "description": "Not trigger: tuning an existing skill (routes to skill-personalizer)",
      "input": "My invoice-parsing skill keeps missing UK VAT fields. Tune it to my company's tooling.",
      "shouldTrigger": false,
      "expect": "Does not mine; defers to skill-personalizer for existing-skill tuning",
      "assertions": [
        "Recognizes this is tuning an existing skill, not mining",
        "Points to skill-personalizer instead of drafting a new 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: 472kUpdated 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: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated 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: 322kUpdated Claude CodeCodexVerifiedInstall