Skip to main content
Agent skill

Build Test

Run the project's build / typecheck / lint / test commands and emit the build.passing + tests.passing signals devloop convergence reads.

Get started

Get the complete skill folder, including its scripts and reference files.

When to use this skill

Use Build Test when an AI agent needs a reusable SKILL.md workflow for this job: Run the project's build / typecheck / lint / test commands and emit the build.passing + tests.passing signals devloop convergence reads.

When not to use it

Skip Build Test when the task is outside the coding category, or when a more specific skill in this directory already covers the same workflow with clearer triggers.

How to install

Where will you use it?

Install scope
Claude Code folder setup
mkdir -p ~/.claude/skills/build-test
# Save SKILL.md to ~/.claude/skills/build-test/SKILL.md
  1. Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/build-test/SKILL.md. For a shared project, commit the folder at .claude/skills/build-test/ instead.
  2. Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
  3. For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.

Full install guide for Claude, Cursor, and Codex →

Try a starter prompt

After installing the skill, paste this starter prompt into your agent to verify it triggers on a small task.

Starter prompt
Use the Build Test skill for this task: Use Build Test when this documented workflow matches the task. Confirm any required inputs or files before making changes.

What supports this recommendation?

70/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.

Documentation
74/100
Practical detail
68/100
Source evidence
63/100
Source signals
72/100
Model compatibility evidence
  • claudeuntested

    No model-specific signal or recorded compatibility test was found.

  • chatgptuntested

    No model-specific signal or recorded compatibility test was found.

  • geminiuntested

    No model-specific signal or recorded compatibility test was found.

  • copilotuntested

    No model-specific signal or recorded compatibility test was found.

  • llamauntested

    No model-specific signal or recorded compatibility test was found.

  • perplexityuntested

    No model-specific signal or recorded compatibility test was found.

  • mistraluntested

    No model-specific signal or recorded compatibility test was found.

  • grokuntested

    No model-specific signal or recorded compatibility test was found.

Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.

What this skill does

# Build test

Spec §20.2 / §22.4: the "subjective `critique-theater` panel" is not enough on its own — a code-migration / tune-collab run must prove the build still passes. This atom shells out to the project's declared test commands and emits structured signals the devloop's `until` evaluator reads.

## Inputs

- The active target stack (recorded in `code/index.json` + `plan.md`'s targetStack block). - The user's optional override (`od plugin run --input testCommand='pnpm test'`).

## Default commands by framework

| framework | typecheck | test | | --- | --- | --- | | next | `pnpm typecheck` | `pnpm test` | | vite | `pnpm typecheck` | `pnpm test` | | remix | `pnpm typecheck` | `pnpm test` | | custom | (read from `package.json.scripts`) | (read from `package.json.scripts`) |

## Output

```text project-cwd/ └── critique/ ├── build-test.json # { build: 'passing' | 'failing', tests: 'passing' | 'failing' | 'skipped', durationMs, commandsRun: [...], failures: [...] } └── build-test.log # raw stdout / stderr (truncated to the runner's logBudgetBytes limit, default 1 MiB) ```

The atom emits **two** signals the devloop reads:

- `build.passing: boolean` — the typecheck command exited 0. - `tests.passing: boolean` — the test command exited 0.

Plus the legacy `critique.score` so existing pipelines that read the score keep working: 5 when both pass, 3 when only build passes, 1 when both fail.

## Convergence

Pipelines wire the atom into a devloop:

```jsonc { "id": "verify", "atoms": ["patch-edit", "build-test"], "repeat": true, "until": "(build.passing && tests.passing) || iterations >= 8" } ```

## Anti-patterns the prompt fragment forbids

- Running tests on a partially-edited working tree (the atom always runs after `patch-edit`, and only when `plan.steps`'s current step is in `completed` state). - Suppressing test failures by changing test files; that violates the rewrite-plan ownership rule. - Setting `tests: 'skipped'` without a non-empty `reason` field.

## Status

Implemented by the daemon runner in `apps/daemon/src/plugins/atoms/build-test.ts`. It executes the configured build and test commands, bounds runtime and captured logs, and persists the report and log outputs above.

Intended uses

  • Use Build Test when this documented workflow matches the task.

Implementation notes

  • Review the source instructions and adapt inputs before running the workflow.

Try it, then tell us how it went

A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.