Agent-design-language repo-review-tests

Specialist test reviewer for a multi-agent repository review. Use when a review packet needs a bounded test-quality role focused on missing coverage, weak assertions, brittle fixtures, validation gaps, test isolation, flaky or overbroad tests, and whether risky behavior has executable proof without editing tests.

install

source · Clone the upstream repo

git clone https://github.com/danielbaustin/agent-design-language

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/danielbaustin/agent-design-language "$T" && mkdir -p ~/.claude/skills && cp -r "$T/adl/tools/skills/repo-review-tests" ~/.claude/skills/danielbaustin-agent-design-language-repo-review-tests && rm -rf "$T"

manifest: adl/tools/skills/repo-review-tests/SKILL.md

source content

Repo Review Tests

Review the test and validation surface as one specialist in the multi-agent repo review suite.

This skill identifies coverage and validation risks. It does not write tests; use

test-generator

later if remediation is requested.

Quick Start

Confirm the repo, branch, path, diff, or review packet scope.
Identify risky behavior from the target and map it to existing tests.
Look for missing coverage, weak assertions, brittle fixtures, and validation commands that overclaim proof.
Emit findings first, with affected behavior and the missing proof.
Hand the artifact to
```
repo-review-synthesis
```
for cross-role assembly.

Focus

Prioritize:

risky code paths without direct tests
weak assertions that only prove the happy path ran
brittle fixtures, hidden ordering assumptions, and flaky timing
tests that encode stale behavior or contradict docs
validation commands that are too broad, too narrow, or misleading
missing negative cases around parsing, permissions, retries, recovery, and IO

Defer primary ownership of these areas to other specialists:

implementation defects:
```
repo-review-code
```
security exploitability:
```
repo-review-security
```
documentation truth:
```
repo-review-docs
```
final dedupe and ordering:
```
repo-review-synthesis
```

Required Inputs

At minimum, gather:

```
repo_root
```

one concrete target:

```
target.target_path
```
```
target.branch
```
```
target.diff_base
```
```
target.review_packet_path
```

Useful additional inputs:

```
changed_paths
```
```
risky_behaviors
```
```
test_commands
```
```
exclude_paths
```
```
validation_mode
```

If there is no concrete repo or slice target, stop and report

blocked

Output Expectations

Default output should include:

findings first
reviewed test surfaces
missing proof map
validation performed or not run
residual test risk

Use the shared suite contract in

adl/tools/skills/docs/MULTI_AGENT_REPO_REVIEW_SKILL_SUITE.md

when ADL expects a structured artifact.

Stop Boundary

Stop after producing the test-review artifact.

Do not:

write tests or fixtures
claim coverage that was not executed
block on exhaustive test strategy when a bounded missing-proof finding is enough
downgrade code or security findings from other specialists