SkillAgentSearch skills...

scenario-unity-expert

Use when an agent drives Unity 6.3 (6000.3) on a Mac for any task: batch mode and -executeMethod jobs, a running editor (the user's GUI editor or a headless one), the Unity CLI and com.unity.pipeline, MCP servers, the Test Framework, command-line builds; when a batch run fails silently, 'Scripts hav…

Install / Use

npx skills add scenario-labs/skills --skill scenario-unity-expert

Installs into whichever agent you are using.

About this skill
📄

SKILL.md

Installable skill definition

Quality Score

86/100

Category

Automation

Supported Platforms

Universal

Our assessment of scenario-unity-expert

scenario-unity-expert scores 86/100 on our quality scale, 1794th of 2,758 Automation skills we index.

Its SKILL.md is 21 KB long, well organised into 13 sections and no code examples: a thorough specification that gives an agent plenty to work with.

With 1,048 GitHub stars, it is one of the more widely adopted skills in the catalogue.

Substance
30/30
Structure
13/20
Description
15/15
Adoption
13/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated 35 days ago, so scenario-unity-expert is actively maintained.
  • It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands.

Automated pattern scan on 2026-11-20. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

scenario-unity-expert compared with similar skills

All 4 of these similar skills score higher than scenario-unity-expert; compare them before choosing.

SkillScoreStarsUpdatedFormat
scenario-unity-expert (this skill)by scenario-labs861.0k35d agoSKILL.md
Agent-Reachby Panniantong100114.0ktodayCLAUDE.md
hyperframesby heygen-com10088.2ktodayCLAUDE.md
rufloby ruvnet10082.3ktodayCLAUDE.md
headroomby headroomlabs-ai10080.5ktodayCLAUDE.md

Frequently asked questions

How do I install scenario-unity-expert?
Run npx skills add scenario-labs/skills --skill scenario-unity-expert. The install tabs above show the steps for each supported agent.
Which AI agents does scenario-unity-expert work with?
It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
Is scenario-unity-expert safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is scenario-unity-expert still maintained?
The repository was last updated 35 days ago, so scenario-unity-expert is actively maintained.

name: scenario-unity-expert description: "Use when an agent drives Unity 6.3 (6000.3) on a Mac for any task: batch mode and -executeMethod jobs, a running editor (the user's GUI editor or a headless one), the Unity CLI and com.unity.pipeline, MCP servers, the Test Framework, command-line builds; when a batch run fails silently, 'Scripts have compiler errors', a capture comes back black or pink, or a result must be proven with numbers and a frame; or when a Unity brief must go to the right scenario-unity-* specialist." license: MIT

Unity expert (technical director and agent protocol)

Expert Unity work is a loop run against numbers: set the budget, build one stage, measure it, look at it, fix it, then advance. An agent without a mouse gets there by choosing the right channel for each action, editing through serialization with explicit saves, and never trusting its own eyes alone. This skill is that protocol, the shared toolkit every scenario-unity-* skill imports (scripts/), and the map of the team. Target: Unity 6000.3.21f1 (6.3 LTS), URP 17.3, macOS Apple Silicon. If a sibling skill named here is missing from your available skills, load it with skill_get (Scenario MCP) or ask the user to install it (https://github.com/scenario-labs/skills#install); unattended, proceed from tool schemas and flag the gap.

Status (2026-09-24, after blind grading and the refactor): every toolkit function ran in Unity 6000.3.21f1 here (tests/code/unity-expert/run_all.sh --live --builds; results in archive/tests/unity-expert/live_results.jsonl). Every call from before the refactor is unchanged.

Stance (the expert delta)

  1. One editor per project, and the live editor wins. "You can't open a project in batch mode while the Editor has the same project open" (6.3 Manual). When an editor holds the project, drive it (ut_live), never hand-edit .unity/.prefab/.asset YAML: wrong fileIDs, invisible until reimport, easy to hit the wrong scene (Unity's unity-cli skill).
  2. A job is only as honest as its result line. Pass on the AGENT_RESULT envelope and the log scan, not on "it exited". A compile error in ANY assembly aborts every batch job with "Scripts have compiler errors" (observed; the live editor's twin is Safe Mode, which also blocks com.unity.pipeline).
  3. Numbers before pixels, milliseconds not fps. 16.67 ms at 60 fps, 33.33 at 30; mobile keeps about 65% for thermals (Unity profiling e-book). Editor Play mode is for iteration; the verdict comes from a development player on the lowest target device.
  4. Look AND check, and know the capture's lies. Every frame goes through ut_review.image_checks and a contact sheet you open. Epic's lighting agent approved an all-white frame (scenario-unreal-expert sources); here batch captures showed first-frame white materials, culled 2D lights, bind poses, no Overlay UI and black metals (observed by the team): anything wrong only in an Editor capture is re-checked in a player first. Agent-built levels come out "90%" done, a shelf in front of a door (Code Monkey, ppFshgFOgXU [00:10:52]).
  5. Fix the platform at launch. Target switches inside a batch job silently do nothing (6.3 Manual); -buildTarget/-activeBuildProfile at launch, one process per target. The editor reopens on the last active platform (observed): pin it.
  6. Named jobs over eval. eval to explore, a named job for anything repeated: "A written command turns that workflow into something deterministic and repeatable" (DgNrgZeJOxQ [00:07:41]); generic run-command tools drift (git-amend, VPjo-M6mPkE [00:03:54]).
  7. Tests are the contract. Give work a failing test and a fence ("It can only touch the card data", DgNrgZeJOxQ [00:05:31]); -runTests never with -quit.
  8. Be honest about visual editors. Shader Graph, VFX Graph, the Animator window, UI Builder, Tile Palette painting and terrain sculpting have no authoring API: use the substitute (HLSL, templates with exposed properties, AnimatorController API, UXML/USS text, TerrainData) or name the human path.

Channels

| Channel | Use for | Cost (measured here) | Rules | | ------------------------------------------------------------------------------- | ----------------------------------------------------------------------- | ----------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | Batch job ut_run.run_method | anything when no editor holds the project | 6 to 13 s per job (editor boot included) | -nographics by default; graphics=True for captures, bakes (never bake with -nographics), profiling; quit=False for async jobs; an editor that hangs after its result is ended by PID (grace, 10 s) | | Resident headless editor + AgentKit bridge ut_live.start_headless, call | many calls in a row | boot 5 to 20 s once, then 0.1 to 0.5 s per call | holds the project and a license seat; stop_headless | | User's GUI editor + AgentKit bridge ut_live.call, eval | the user has the project open | same code path as headless (not run: no GUI windows during tests) | the user may need to focus Unity once so it compiles AgentKit; never re-open the user's scene without asking; save what you change | | Unity CLI + com.unity.pipeline ut_live.cli_*, eval (auto) | fastest live channel: 151 commands, Roslyn eval with no domain reload | eval 0.03 to 0.15 s after a 0.5 to 1.1 s first call | CLI 1.0.0-beta.11 (portable copy), package 0.7.0-exp.1 installed before the editor opens; blocked by Safe Mode, possibly by a sandbox | | MCP | only unity mcp or a vetted community server | | Unity's AI Assistant MCP server is deprecated ("Use the Unity CLI instead") | | Test Framework ut_run.run_tests | EditMode and PlayMode verdicts, content validation | about 14 s per platform | NUnit XML parsed; exit 2 when a test failed (observed) |

ut_live.channel(P) picks batch, bridge, pipeline, or blocked (ask the user). "Can't connect" does not prove the editor is closed: rule out Safe Mode and a sandboxed shell first (unity-cli skill).

Establish first

Ask once, then run: target platforms and device class; frame budget (default 60 fps desktop, 30 fps mobile at 65%); 2D or 3D; pipeline (default URP 17.3, HDRP only for high-end PC or console); quality tiers; art style; whether the user's editor has the project open; deliverable. Then ut_env.preflight(P) and AgentJob.Echo; write what they report (pipeline asset, color space, active target, quality level, Enter Play Mode setting, input handler, scripting backend) into the plan.

Workflow (the expert loop)

  1. Brief in numbers. Budgets per platform, resolution, scale, style. GATE: every later check has a number to compare with.
  2. Preflight. License days left, modules, lock state, compile health (one Echo job). GATE: ok, zero compile errors, zero new CS0618 warnings.
  3. Baseline. AgentAudit.AuditScene/AuditAssets; captures of fixed bookmarks (AgentCapture.SaveBookmark); a Play-mode profile if performance matters. GATE: audit errors fixed or listed; contact sheet opened.
  4. Build one stage with small idempotent jobs (references/writing-jobs.md): get-or-create, serialized edits, explicit saves. Route domain work to the owner (below).
  5. Measure. Re-run the audit, the tests, AgentProfile.PlayModeTimings + ut_stat.budget_check, or the build summary. GATE: numbers against the brief.
  6. Look AND check. Same bookmarks, ut_review.review_capture, open the sheet, judge with the domain skill's critique. GATE: no blank or magenta frame, saturation not collapsed, the change visible.
  7. Fix before advancing, one change at a time; ut_review.compare and ut_stat.compare_frames prove parity or gain.
  8. Deliver with evidence. Each step with its Unity call (references/toolkit-map.md), then Verified (what ran, job id, number or frame) and Assumed (not run, [verify], defaults). Never call a scene, shader, level or build done without numbers and a frame you looked at.

Numbers

| Value | Relative to | Source | | --------------------------------- | ------------------------------------------------------------------------------------------ | ---------------------------- | | 16.67 / 33.33 ms | frame budget at 60 / 30 fps | profiling e-book | | x 0.65 | mobile sustained budget (10.8 ms at 60, 21.7 at 30) | profiling and mobile e-books | | 6 to 13 s | one batch job on this project, editor boot included | observed | | about 14 s | one run_tests platform run (EditMode 2 tests, PlayMode 1 test) | observed | | 60 s (12 s incremental), 114.7 MB | macOS Mono player build of the demo scene (universal) | observed | | 259 s (25 s incremental), 12.2 MB | Web build (IL2CPP, Brotli) of the same scene | observed | | 0.1 to 0.5 s / 0.03 to 0.15 s | live call through the AgentKit bridge / Pipeline eval (two runs) | observed | | 30 days | Personal license offline window (license_ok() reads it) | observed + Unity support FAQ | | 10 s / 120 s | grace before ut_run ends a silent editor after its result with / without a shutdown line | toolkit default |

Quality gates

  • Measurable: envelope ok plus compile_errors == []; audit counts.error == 0; NUnit failed == 0 and total > 0; budget_check verdict pass (warn needs a reason); `gc_ch

Truncated for display — read the full file on GitHub.

Related Skills

View on GitHub
GitHub Stars1.0k
CategoryAutomation
Updated1mo ago
Forks129

Languages

Python

Trust signals

100/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

No cautions