legal-ai-engineering-lab / tests
AGENTS.md2 starsChanged 3 months agoMITView on GitHub
These rules apply when writing or changing tests. AGENTS.md at the repository root governs everything else. - Use a test wherever an exact assertion is possible — date parsing, validation rules, permission checks, state transitions, duplicate prevention, schema enforcement. Do not reach for an AI evaluation where an ordinary assertion answers the question precisely; evaluations are slower, noisier, and weaker evidence when the answer is deterministic. - The expected value must come from a source independent of the code being…
In this repository21 other files it gives its agents.
- .agents/AGENTS.mdAGENTS.md
- AGENTS.mdAGENTS.md
- .claude/rules/agents.mdAGENTS.md
- evals/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- architecture-reviewSkill
- code-reviewSkill
- evaluation-reviewSkill
- explainSkill
- interview-reviewSkill
- learning-reviewSkill
- security-reviewSkill
- start-projectSkill
- architecture-reviewSkill
- code-reviewSkill
- evaluation-reviewSkill
- explainSkill
- interview-reviewSkill
- learning-reviewSkill
- security-reviewSkill
- start-projectSkill
The file
Files
<!-- claude-rule-paths: tests/** -->Working in tests/
These rules apply when writing or changing tests. AGENTS.md at the repository root governs
everything else.
- Use a test wherever an exact assertion is possible — date parsing, validation rules, permission checks, state transitions, duplicate prevention, schema enforcement. Do not reach for an AI evaluation where an ordinary assertion answers the question precisely; evaluations are slower, noisier, and weaker evidence when the answer is deterministic.
- The expected value must come from a source independent of the code being tested. A test whose expectation is calculated the way the code calculates it cannot fail for the right reason.
- This directory covers
src/legal_lab/and the repository's own tooling, not experiments. Experiments carry their own checks locally until promoted; holding exploratory code to production standards is the failure this repository exists to avoid. The one exception is the reference experiment, whose measured figures are quoted as fact in its README — documentation that can go silently stale is worth an assertion. - When fixing a bug that could recur, add the failing case before fixing it. That is the cheapest regression protection, and it makes the fix demonstrable rather than asserted.
- Some tests here protect the instruction system itself — that references resolve, that skills carry triggers, that the always-on context stays within budget. Treat those as load-bearing: the agent's behaviour is part of this repository's product.
tests/README.md carries the division of labour between tests and evaluations.
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
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.

