Help us improve
Share bugs, ideas, or general feedback.
From lumina
Compile the Unity project headlessly, run play-mode/edit-mode tests, and return structured compile/runtime errors for the self-correct loop. Use after writing or changing scripts, before declaring anything done.
npx claudepluginhub tapascodespace/claude-plugin --plugin luminaHow this skill is triggered — by the user, by Claude, or both
Slash command
/lumina:build-verifyThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Turn "I wrote some code" into "it actually compiles and runs." This is the verifier half of the
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Guides systematic root-cause debugging when tests fail, builds break, or unexpected errors occur. Provides a structured triage checklist to preserve evidence, localize, and fix issues instead of guessing.
Share bugs, ideas, or general feedback.
Turn "I wrote some code" into "it actually compiles and runs." This is the verifier half of the loop. Return structured, actionable errors so the orchestrator can self-correct.
Project path (default = current dir): $ARGUMENTS
Two complementary checks — do the cheap one first:
Compile check (fast). Trigger a script recompile and read the result.
mcp__unity) to recompile and return console errors — this is
fastest because the Editor is already running.Unity -batchmode -quit -projectPath "<path>" -logFile -
Capture stdout/stderr; the compile errors appear in the log.Tests (if present). Run Unity Test Runner headlessly:
Unity -batchmode -runTests -projectPath "<path>" -testPlatform PlayMode -testResults "<path>/TestResults.xml" -logFile -
Also run -testPlatform EditMode if edit-mode tests exist. Parse TestResults.xml.
Always return a compact, structured summary the orchestrator can act on:
RESULT: pass | fail
COMPILE:
- <File.cs>:<line> <CSxxxx> <message>
TESTS: <n passed, m failed>
- FAILED <TestName>: <assertion / message>
NEXT: <one-line hypothesis for the most likely root cause, if failed>
TestResults.xml.