20 Claude Code agents. Paste one in. Ship real work.

A library of tested agent blueprints — PR reviewer, migration planner, bug triage, release notes, incident post-mortems, 15 more. Every agent comes with the prompts, the skill files, and a sample run showing what it produces.

$129 at launchRetail $179Waitlist = 25% off launch week

Join the waitlist

Get the launch email

Waitlist members get 25% off the first 48 hours after launch, automatically. If an agent you want isn't in the list below, reply to your confirmation email and I'll add it if there's demand.

What's in the library

20 agents. One file each. Full source.

pr-reviewer

Full-PR review that checks tests, docs, type changes, and rollback safety.

test-backfill

Adds tests to untested modules using the project's own test conventions.

migration-planner

Scans for deprecated API usage and sequences the migration into safe commits.

bug-triage

Reads a bug report + stack trace, returns a repro plan + suspect files.

release-notes

Generates human-readable release notes from commits since last tag.

incident-postmortem

Drafts an incident report from Slack / commit history / logs.

dep-upgrade

Bumps a dependency, fixes the breakage, and writes the PR description.

security-sweep

Runs a focused sweep for the OWASP top-10 patterns in a given directory.

docs-sync

Keeps README examples in sync with the actual API by running them.

backfill-types

Adds TypeScript types to a JavaScript module without changing runtime behavior.

i18n-extractor

Finds hardcoded strings and extracts them to the i18n catalog.

flaky-hunter

Identifies flaky tests from CI history and drafts fixes.

db-migration-author

Writes reversible DB migrations with a dry-run checker.

api-contract-diff

Compares API schemas between two branches and flags breaking changes.

onboarding-map

Generates an onboarding tour of a codebase for a new hire.

dead-code

Finds genuinely dead code (call-graph reachability, not regex).

perf-profile-reader

Reads a flame graph or pprof output and names the real bottleneck.

config-audit

Scans config files for secrets, drift, and unreferenced keys.

slo-drafter

Writes an SLO proposal given a service's dashboards and latency history.

prompt-test-runner

Runs an A/B test on two prompts across a fixed eval set and reports deltas.