All loops
Claude CodeMedium 7 min· claude-opus-4-1
Test gap finder
Comments on PRs with the 3 highest-risk untested branches.
NOT DEPLOYEDNOT DEPLOYED
0142ms
Trigger
POST https://hooks.claudeloops.run/test-gap-finder · event=slack.slash:/loop
02922ms
Agent
claude-opus-4-1 · in 822 tok · out 545 tok
03212ms
Tools
github-mcp/github:pulls.create → 200 OK · 262ms
0432ms
Verify
schema check · zod v3 passed
0562ms
Output
changelog updated · v1.4.2
0622ms
Notify
audit log written · runbook link attached
SUCCESS
0%
0 runs
P50
0ms
median
P95
0ms
tail
AVG COST
—
per run
LAST OK
never
—
LAST FAIL
never
none
Latency · last 30 runs0 samples
no runs yet
Latest output · what your users see
Openacme/infra #1842
feat: add /health endpoint
5 files+202-12
- ✓ Ran full test suite → 218 passed, 0 failed (12.3s)
- ✓ Coverage delta: +0.4% on `src/lib/*`
- ✓ No breaking changes detected in public API
// press Test to run once · Watch live to keep streaming · Deploy to make it real
The problem
github,tests
The outcome
Coverage tools tell you a number; they don't tell you which untested branch will blow up.
Ingredients & skills
Secrets
- ANTHROPIC_API_KEY
- GITHUB_TOKEN
Providers
- Anthropic
- GitHub
MCP servers
- github-mcp
#claude-code#github#automation
How it works
Comments on PRs with the 3 highest-risk untested branches.
Step 1
1 — Create the locker
Locker `test-gap-finder` holds the GitHub token (repo scope) and the Anthropic key.
bash
locker create test-gap-finder
locker set $_ ANTHROPIC_API_KEY=sk-ant-...
locker set $_ GITHUB_TOKEN=ghp_...Step 2
2 — Wire the GitHub event
Subscribe to the relevant GitHub webhook through the `github-mcp` server. Claude only sees the event payload and the files it requests.
Step 3
3 — Ship the agent
Single TS file in `agents/`. Stateless; the locker is bound at runtime so no env editing for new repos.
One-line deploy
The button above runs the same command with your saved config. This is the raw CLI form.
bash
npx claudeloops deploy test-gap-finder https://hooks.claudeloops.run/test-gap-finder
Related loops
Claude Code
PR review → risk-tagged comment
Claude Code reviews every PR and posts a risk-banded comment with suggested tests.
Claude Code
Test synthesizer for changed files
Claude Code writes Vitest + Playwright tests for every file touched in a PR.
Claude Code
Schema diff → migrations bot
Claude Code reads schema.prisma diffs, writes safe SQL with rollback notes.