justnnc/justnnc-skills

{
  "$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
  "name": "justnnc-skills",
  "version": "0.1.0",
  "description": "Agent skills for software work that has to be right, not just plausible: prove a tool on a contained copy, fan out a large refactor safely, audit specs against what landed, take a feature idea to a decided design, probe how a codebase works, attack a proposal, and write better skills.",
  "author": { "name": "justnnc", "url": "https://github.com/justnnc" },
  "homepage": "https://github.com/justnnc/skills",
  "repository": "https://github.com/justnnc/skills",
  "license": "MIT",
  "keywords": ["skills", "verification", "refactoring", "specs", "code-review", "design"],
  "extensions": {
    "com.openai": {
      "interface": {
        "displayName": "justnnc skills",
        "shortDescription": "Measure twice, cut once",
        "longDescription": "Agent skills for software work that has to be right, not just plausible. Each one asks the agent to show its evidence: commands it ran, lines it read, decisions you made.\n\n- prove-by-running: runs a script or tool on a disposable, contained copy and ends in one verdict: runs clean, breaks, or unproven.\n- parallel-refactor: fans one large mechanical change out across parallel agents: measure the real call sites, write a contract, keep every intermediate state valid, verify with a canary.\n- spec-audit: audits specs against the work that just landed: built but undescribed, described but unbuilt, stale text, planned text inside a spec.\n- feature-spec: takes a feature idea to a decided design document: what already exists, a security review of candidates, quick spikes of the risky parts, each decision recorded.\n- how-it-works: answers how a codebase works in two checked stages: one maps where to look, the next reads those spots closely.\n- adversarial-review: briefs a reviewer to attack a proposal before it is built: failure modes, weak evidence, simpler alternatives.\n- skill-craft: a checklist for writing, reviewing and testing agent skills.\n\nWritten for engineers who direct coding agents and want careful, verifiable work over fast, plausible work.\n\nLimitations: the plugin contains only markdown skill files, with no scripts, hooks, MCP servers or network calls. Most skills direct an agent that can run commands, use git and start subagents; where it cannot, those steps become instructions for you. The skills are new and have not been evaluated across models, so treat them as considered practice rather than measured results. Source and issues: https://github.com/justnnc/skills",
        "developerName": "justnnc",
        "category": "Developer Tools",
        "capabilities": [],
        "supportURL": "https://github.com/justnnc/skills/issues",
        "privacyPolicyURL": "https://github.com/justnnc/skills/blob/main/PRIVACY.md",
        "composerIcon": "./assets/icon.png",
        "logo": "./assets/icon.png"
      },
      "publication": {
        "release_notes": "First release: seven skills (prove-by-running, parallel-refactor, spec-audit, feature-spec, how-it-works, adversarial-review, skill-craft)."
      }
    }
  }
}