agentleFS
45ck45ck/video-evaluator/skills/review-bundle/SKILL.md

review-bundle

Skill2 starsChanged 7 months agoMITView on GitHub

Review a local video artifact bundle, summarize report status across known artifacts, and point the agent at the first useful review surfaces.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 1 is distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this when a repo already produced a run and the agent needs the shared review layer rather than repo-specific generation logic.

    line 8
  2. Very specific:

    Use the same bundle locators as video-artifact-intake, plus optional prompt hints:

    line 19
  3. Very specific:

    Run video-artifact-intake first only if path resolution is ambiguous.

    line 40
  4. Very specific:

    Treat fail or warn as routing signals to inspect the named artifact, not as complete diagnostics.

    line 49
  5. Very specific:

    Do not claim the video is correct because overallStatus is pass.

    line 56
  6. Very specific:

    If the run directory is wrong, the review will faithfully summarize the wrong bundle.

    line 62

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Review Bundle

Use this when a repo already produced a run and the agent needs the shared review layer rather than repo-specific generation logic.

This skill does not replace domain-specific QA. It tells the agent:

  • what artifacts exist
  • which reports passed, warned, or failed
  • which surfaces deserve immediate human or agent attention

Inputs

Use the same bundle locators as video-artifact-intake, plus optional prompt hints:

{
  "outputDir": "./output/example-run",
  "includePromptHints": true
}

includePromptHints defaults to true. Set it to false when you only need a machine-readable bundle/status summary.

Outputs

The result includes the normalized intake result plus a review-oriented summary of artifact availability, report status, recommended focus, and optional prompt hints. It does not open or deeply evaluate every report.

Sequencing Guidance

Run video-artifact-intake first only if path resolution is ambiguous; otherwise review-bundle can do intake itself. If the review identifies storyboard files, inspect them directly or continue with storyboard-transitions/storyboard-understand. If storyboard files are absent and a primary video exists, generate them with storyboard-extract and storyboard-ocr.

Interpretation Notes

Treat fail or warn as routing signals to inspect the named artifact, not as complete diagnostics. Treat unknown as insufficient structured evidence. A present trace.zip means failure evidence may exist outside JSON and should be unpacked or inspected if relevant to the task.

Abstention Rules

Do not claim the video is correct because overallStatus is pass; the skill only summarizes known reports. Abstain from visual, caption, or UX judgments unless the result points to concrete artifacts you inspect.

Failure Modes

If the run directory is wrong, the review will faithfully summarize the wrong bundle. If reports use custom schemas without status, passed, or hasFailures, they may appear only as present; read them manually before concluding they passed.

Repo-side runner:

node --import tsx scripts/harness/review-bundle.ts

Installed-pack runner:

node ./.video-evaluator/agent/run-tool.mjs review-bundle

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.