Skip to content

JPeetz/agent-skills / skills/development/finding-unknowns/skills/blindspot-pass

blindspot-pass

Verified

Use when entering unfamiliar codebase or domain territory.

FileKindSize
CHANGELOG.mdmarkdown389 B
SKILL.mdskill md2.0 KB
evals/evals.jsonconfig2.4 KB

evals/evals.json

View on GitHub
{
  "skill": "blindspot-pass",
  "version": "1.0.0",
  "cases": [
    {
      "id": "e1",
      "description": "Should trigger: unfamiliar codebase territory",
      "input": "I'm about to add authentication to our Express API but I've never worked on it before.",
      "shouldTrigger": true,
      "expect": "Runs a blindspot pass: landmines, hidden context, what good looks like, questions to ask, then rewrites the request",
      "assertions": [
        "Reports landmines first (common mistakes, repo potholes)",
        "Surfaces hidden context and invariants",
        "Shows 2-3 examples of good patterns",
        "Lists 3-5 questions an expert would ask",
        "Ends with rewritten request",
        "Does NOT start implementing"
      ]
    },
    {
      "id": "e2",
      "description": "Should trigger: unfamiliar domain territory",
      "input": "We're taking over maintenance of this payments microservice. What am I missing?",
      "shouldTrigger": true,
      "expect": "Explores the module/history/conventions and reports blindspots",
      "assertions": [
        "Prioritizes unknowns that change architecture over trivia",
        "Plainly says 'no significant blindspots' if that's the case"
      ]
    },
    {
      "id": "n1",
      "description": "Not trigger: familiar territory, direct fix request",
      "input": "Fix the timeout bug in the module I wrote last week.",
      "shouldTrigger": false,
      "expect": "Goes straight to debugging",
      "assertions": [
        "No blindspot report produced"
      ]
    },
    {
      "id": "n2",
      "description": "Not trigger: general request unrelated to entering new territory",
      "input": "Order pizza for the team.",
      "shouldTrigger": false,
      "expect": "Does not engage blindspot workflow",
      "assertions": [
        "No exploration performed"
      ]
    },
    {
      "id": "e3",
      "description": "Should trigger: external-domain blindspots",
      "input": "We're building a compliance feature and I've never worked with GDPR. What are the table stakes?",
      "shouldTrigger": true,
      "expect": "Researches what practitioners consider table stakes and reports landmines/hidden context",
      "assertions": [
        "Covers what practitioners call table stakes in the domain",
        "Reports landmines and hidden context",
        "Does not start implementing",
        "Ends with a rewritten request"
      ]
    }
  ]
}
receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementationStars: 322kUpdated Claude CodeCodexVerifiedInstall
requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirementsStars: 322kUpdated Claude CodeCodexVerifiedInstall
using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallbackStars: 322kUpdated Claude CodeCodexVerifiedInstall
verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions alwaysStars: 322kUpdated Claude CodeCodexVerifiedInstall