Skip to content

TeamRetroHQ/teamretro-skills / skills/ai-retro-brief

ai-retro-brief

Verified

Use when the user asks to synthesize, summarize, or analyze the AI session retro log — recurring friction in AI-assisted work, patterns across AI sessions, or prep for a team retrospective (e.g. "what patterns are in our AI retros", "prep the AI retro summary for Friday"). Companion to ai-session-retro, which writes the entries this skill reads.

evals/evals.json

View on GitHub
{
  "skill_name": "ai-retro-brief",
  "evals": [
    {
      "id": 0,
      "eval_name": "synthesize-brief-from-entries",
      "prompt": "We've been logging AI session retros for a few weeks now (docs/ai-retros/entries/ in this repo). Can you synthesize them into a one-page brief we can bring into our team retrospective on Friday?",
      "expected_output": "Produces docs/ai-retros/briefs/brief-<today>.md with: recurring-friction themes ranked by frequency x cost (ambiguous-instruction should rank #1 at 4 of 6 sessions), root-cause distribution grouped into briefing/documentation/codebase/tooling/agent with briefing dominating, adoption tracking (time-window fix adopted by 2026-07-09; live-search verification not verifiable from the log), top 3 evidence-tied actions citing dated entries, and a trend paragraph honest about the 6-entry sample. Stamped ai-retro-brief v1.2.",
      "files": ["fixtures/six-entries/docs/ai-retros/entries/"]
    },
    {
      "id": 1,
      "eval_name": "too-few-entries-guard",
      "prompt": "Synthesize our AI retro log (docs/ai-retros/entries/) into a brief for tomorrow's retrospective.",
      "expected_output": "With only 2 entries present, states that fewer than 3 entries exist, gives a light summary only, and makes NO pattern or trend claims. Does not fabricate recurrence.",
      "files": ["fixtures/two-entries/docs/ai-retros/entries/2026-07-06-support-ticket-theme-analysis.md", "fixtures/two-entries/docs/ai-retros/entries/2026-07-09-onboarding-email-sequence.md"]
    }
  ]
}
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: 323kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 323kUpdated 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: 323kUpdated Claude CodeCodexVerifiedInstall