AGENT SKILL

tk-studio-playable-judge

The playable-verify leg's screenshot judge (Story 34.3) — reads only a playable run's report.json and its frames, answers the fixed rubric (scene loaded, player visible and framed, no magenta or error overlay, HUD reads, the named moments shown) with pass or fail and findings each naming a frame and a rubric line. Driven headless by the supervisor's host leg on the project's sandbox tier; use when the payload names a playable `media` folder, or the user says "judge the playable frames".

License 许可
MIT
最近更新
2026-11-10
Stars
★ 0
Forks
0

01 前言 Front Matter

name
tk-studio-playable-judge
description
The playable-verify leg's screenshot judge (Story 34.3) — reads only a playable run's report.json and its frames, answers the fixed rubric (scene loaded, player visible and framed, no magenta or error overlay, HUD reads, the named moments shown) with pass or fail and findings each naming a frame and a rubric line. Driven headless by the supervisor's host leg on the project's sandbox tier; use when the payload names a playable `media` folder, or the user says "judge the playable frames".

02 资源文件 1

03 SKILL.md 正文

tk-studio-playable-judge

The host leg's third step (after a green engine gate) runs playable.py run on the host and copies its media into the worktree's ignored .bmad-loop/runs/<engine>/playable/. This session judges those frames. It is a screenshot judge, nothing more.

What it may read

  • Only report.json and the frames (frames/*.png) under the payload's media folder (relative to the payload directory).
  • Never the project's code, scenes, scripts, the spec, the story, the plan or any other file. Do not list, grep or open anything outside media. A judge that read the code would grade intent, not what the game showed.

Behavior (headless)

  1. Prepare through the deterministic core (it validates the report and the frames, and prints the frame list, the moments and the rubric):

    uv run "${CLAUDE_PLUGIN_ROOT}/lib/playable_judge.py" prepare --media "<directory>/<media>"
    

    A blocked result (no report.json, malformed report: <field>, no frames) ends the run blocked with that reason.

  2. Look at every frame the core listed (Read on each PNG) and answer each rubric line by its fixed id:

    id rubric line
    scene-loaded the scene loaded
    player-visible the player is visible and framed
    no-magenta-or-error-overlay no missing-texture magenta and no error overlay
    hud-reads the HUD reads
    moments-shown the scenario's named moments appear in the frames they name
  3. Write the answer to a file under media (answer.json):

    {"verdict": "fail", "findings": [{"frame": "frames/tick-000030.png", "rubric": "hud-reads", "note": "the HUD is clipped off the right edge"}]}
    

    verdict is pass or fail. Every finding names one frame from the core's list and one rubric id, with a one-line note. A fail carries at least one finding. A pass may carry none.

  4. Record the verdict through the core (it writes <media>/judge.json, the file the host leg reads back):

    uv run "${CLAUDE_PLUGIN_ROOT}/lib/playable_judge.py" verdict --media "<directory>/<media>" --answer "<directory>/<media>/answer.json"
    

    A refusal (unknown rubric id, frame not in frames[], fail with no findings, malformed answer) may be corrected once by rewriting the answer; a second refusal ends the run blocked with that reason.

  5. End with the status block (the verdict lives in judge.json, never in the block; a judged fail is still a complete run):

    {"status": "complete", "intent": "tk-studio-playable-judge", "artifacts": ["judge.json"], "reason": null}
    

If the deterministic core is unrunnable — a tool call denied by permissions, uv/python unavailable — end blocked with the status block naming the unrunnable core as reason: never a question, never a headless run that ends without the block (AD-11).

Rules

  • Never prompt or ask: everything arrives in the payload (media, directory); a missing media is a blocked status (AD-11).
  • Never edit the frames, report.json or anything outside media; the only files this session writes are answer.json and (through the core) judge.json.
  • All plugin paths resolve through ${CLAUDE_PLUGIN_ROOT}.

04 同仓库更多 Skills

thkennedy
thkennedy/tk-studio

tk-studio-track

Tracking verbs on the bound Jira project — `block` moves a story the loop cannot progress, and each dependent story, to "Blocked - Awaiting User" with a comment that @mentions the owner (or, with `--engagement`, one engagement's questions issue, which `close` later moves to "Done"); `reply` posts the owner's answer (or "Done.") with the owner's host-only token and moves the issue to "Block Resolved"; `validate` checks that resolution and passes the issue back to work or re-blocks it with a follow-up. Use when the user says "block this story", "tk track block", "put it in the owner's inbox", "reply to the blocked issue", "validate the resolution", or a supervisor or engagement payload names this skill.

★ 0
thkennedy
thkennedy/tk-studio

tk-studio-session

Session discipline as a surface — land a boundary handoff (optionally carrying knowledge deltas) and end the session, or resume a fresh session from a run workspace alone. Use when the user says "hand off this run", "tk session", "resume the run", or a boundary/budget trigger says the session must split.

★ 0
thkennedy
thkennedy/tk-studio

tk-studio-research

Run a chartered research pass — survey the topics a charter scopes, grade every finding by evidence, land the findings in the run workspace, and emit observation events for recommendation-worthy ones. Use when the user says "run the ecosystem watch", "tk research", or a research job's invoke-skill directive names this skill.

★ 0
thkennedy
thkennedy/tk-studio

tk-studio-report

One-verb human defect report — "the skill did the wrong thing" — recorded as a report event in the local measurement ledger. Use when the user says "tk report", "report a defect", "that skill misbehaved", or "log a studio issue".

★ 0
thkennedy
thkennedy/tk-studio

tk-studio-plan-sync

Planning adapter — canonicalize BMad-native planning artifacts (normalize pass, sole id authority) and sync them with the bound planning backend. Use when the user says "sync the plan", "normalize planning artifacts", "tk plan sync", or after stock BMad planning flows produce or change epics/stories.

★ 0
thkennedy
thkennedy/tk-studio

tk-studio-orchestrator

The studio's front door — resolve who you are (role, per-user store), what this project needs (confirmed working_set.<role>), and route to resources by name, statelessly. Use when the user says "enter the studio", "tk studio", "orchestrate", "convene the council", or asks what to work on through the studio.

★ 0