Which AI agents does this loop work with?

Ralph a test backlog is written for claude-code, codex.

Is this loop safe to run unattended?

This loop is graded A (95/100) by looprepo's automatic safety evaluator, which checks for exit conditions, iteration budgets, verification steps, and scoping before publishing.

Loop/ralph Testinghigh riskadvancedsafety A · 95 · open-ended Geoffrey Huntley

Ralph a test backlog

Iterate over a prioritized list of untested modules with fresh context each pass, writing real behavioral tests for one module at a time and banking lessons in a guardrails file.

prompt

→ Claude Code

/loop each iteration with fresh context: read .ralph/test-backlog.json and .ralph/guardrails.md, pick the top unfinished module, write behavioral tests for its public API (no snapshot-only tests), run the suite, and mark the module done only when its tests pass and coverage for it exceeds 80%; append any discovered testing gotcha (fixtures, mocking rules, async traps) to .ralph/guardrails.md; stop when the backlog is empty or after 25 turns

claude-code · codex

Implementation note

Generate the backlog from your coverage report, worst-covered first. The guardrails file quickly accumulates project-specific mocking rules that make later iterations noticeably better.

Source: Geoffrey Huntley (Ralph technique) ↗graded A · 95/100 — how grades work →

More testing loops

Build a REST API with tests

Loop/ralphGitHub B

Run autonomous iterations to ship a complete REST API with full test coverage until completion is promised.

prompt

→ Claude Code

/ralph-loop "Build a REST API with tests" --max-iterations 30 --completion-promise "COMPLETE

testingmedium risk

claude-code

MaouStan ↗

Test all endpoints

Loop/goalGitHub B

Run tests against every API endpoint until coverage is complete or you hit 30 iterations.

prompt

→ Claude Code

/goal all endpoints tested or stop after 30 turns

testingmedium risk

claude-code

Sol0-dev ↗

Hit acceptance criteria

Loop/goalX B

Drive a feature to done against explicit acceptance criteria: a working paginated endpoint, passing tests, clean lint, and a hard turn cap.

prompt

→ Claude Code

/goal the /users endpoint returns 200 with a paginated JSON body, all tests pass, and lint is clean — stop after 20 turns

testingmedium risk

claude-code codex

akshay_pachaar ↗