agentleFS
witnesstodarkwitnesstodark/mr-mak-workspace/.claude/skills/gameplay-visual-review/SKILL.md

gameplay-visual-review

Skill800 starsChanged 23 days agoMITView on GitHub

Verify a visual game change using controlled native captures, targeted behavioral checks and a reviewable evidence record. Use for animation, VFX, HUD or scene changes whose appearance must be judged in the actual game.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the requested change and its accepted baseline.

    line 8
  2. Distinctive:

    Use the game's actual camera, lighting, materials and normal playback speed.

    line 15
  3. Distinctive:

    Keep seed, subjects, route/action and framing comparable for OFF/ON or before/after captures.

    line 16
  4. Distinctive:

    If synthetic input is needed, separate it from physical devices and restore the prior input settings afterward.

    line 22

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Gameplay visual review

Read the requested change and its accepted baseline. Choose a short test that can reveal the difference. Preserve the current scene and unrelated systems; inspect editor state before taking control, and coordinate if another agent is already using the same editor. Do not restart an active tool merely for a capture.

Build a fair comparison

Use the game's actual camera, lighting, materials and normal playback speed. Keep seed, subjects, route/action and framing comparable for OFF/ON or before/after captures. Record any unavoidable difference. A slow-motion close-up is useful diagnostic evidence, but it should not replace normal gameplay zoom.

A review lab should isolate the changed feature while retaining its production code path. Do not "fix" a comparison with a separate demonstration implementation. If synthetic input is needed, separate it from physical devices and restore the prior input settings afterward. Discard transient test orders rather than saving them into the authored scene.

Check appearance and behavior separately

For animation, inspect contact, seams, interrupted transitions and responsive controls. For VFX, inspect contrast, overlap, attachment and cleanup. For UI, inspect readability, overflow, input blocking and real state. For terrain, inspect scale, grounding, traversal and the gameplay camera.

Counters, hashes and saved values can prove wiring or unchanged inputs. They do not prove visual quality. A screenshot does not prove a timed transition, and a passing browser study does not prove native rendering cost. Measure the relevant system only, and keep unrelated sound/narration quiet during visual-only checks.

Leave useful evidence

Record the source revision, saved configuration, exact scenario, capture type, observations and remaining issues. Label generated art, browser schematics, isolated engine previews and native gameplay clearly. A successful agent check does not imply the user accepted the result.

Provide short local loops or videos for motion and stills for detail. Keep the original capture and a lightweight review copy with provenance. Do not repeatedly recompress an already compressed delivery. Update the current specification when behavior changes; put chronological details in the development log.

Example: compare a shield impact OFF/ON on dark and pale ground with identical gameplay. Deliver the visible comparison plus cancellation and cleanup results; leave final aesthetic acceptance explicit. Dependencies: native engine access and a local capture method; a report viewer is optional.

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.