The library

Everything we index — ranked by what works, never by stars.

untested
Block destructive commands outside projectspluginEngineeringOpsL2
project-boundary · Protecting system files when an LLM agent has shell access
untested
Build and publish Claude plugins in 23 stepspluginEngineeringL3
public-plugin-builder · Rapid prototyping of single-skill plugins where 23-step guided interview ensures correct scoping
untested
Multi-tier code review automationpluginEngineeringL2
ralph-review-trio · Comprehensive solo code reviews where progressive model escalation catches breadth then depth
untested
Generate AI images video music audiopluginMarketingProductL2
runapi-mcp · Single unified interface for multi-provider generative content when provider-switching overhead matters
untested
Catch and fix AI-generated code defectspluginEngineeringL2
slopmop · CI gates preventing AI-generated code quality regressions before merge
untested
Approve or block AI coding decisionspluginEngineeringL2
thumbgate · Learning guardrails from real agent failures in production with NIST/SOC2 audit trail
untested
Audit ship and scale production codepluginEngineeringL3
ultraship · Production-grade pre-deploy gate closing 29 audit dimensions in one step
untested
Screen startups and model financialspluginFinanceL3
venture-capital-intelligence · Rapid bulk-deal screening where 9 specialized skills replace manual analysis across screening/modeling/reports
untested
Research companies and enrich leadspluginSalesMarketingL2
vibe-prospecting · Personalized outreach campaigns where live B2B data and tech-stack intel reduce cold-call blindness
untested
Scan fix and score security vulnerabilitiespluginEngineeringL2
vulnetix · Translating vulnerability intelligence into prioritized remediation where CVSS/EPSS guides triage
untested
Manage ads SEO and site optimizationpluginMarketingL3
notfair · Unified SEO/SEM/GEO workflow where 18 skills cover research, build, optimize, and monitor across Google/Meta
untested
Self-correcting multi-agent coding workflowpluginEngineeringL4
pro-workflow · Structured external red-team assessments where 71 skills auto-select by vulnerability class
untested
Optimize site performance and accessibilitypluginEngineeringProductL2
web-quality-skills · Complex interactive artifacts requiring state management or multi-page routing beyond static HTML
untested
Execute SEO and content marketing at scalepluginMarketingL3
aaron-seo-geo · Full-cycle SEO/GEO where 20 skills + shared contract cover research/build/optimize/monitor with entity truth
untested
Automate security testing and red-teamingpluginEngineeringL4
claude-bughunter · Long-running AI coding projects where persistent memory and cross-agent coordination prevent forgetting context
untested
Build complex interactive web componentspluginEngineeringL2
artifacts-builder · Continuous quality gates where Lighthouse + WCAG + SEO audits run pre-deploy to prevent regression
untested
Multi-agent iterative code reviewpluginEngineeringL3
audit-project · Enforcing zero-defect standards on medium codebases where iterative polish beats single-pass review.
untested
Streamline bug fixing and testingpluginEngineeringL2
bug-fix · Automating the entire bug-to-merge workflow when formalized issue tracking and branch discipline matter more than speed.
untested
Auto-generate customer-friendly release notespluginEngineeringProductL2
changelog-generator · Translating technical git logs into polished release notes when manual changelog writing is too slow or inconsistent.
untested
Create conventional git commitspluginEngineeringL2
commit · Standardizing commit messages across team when manual formatting is error-prone or slow.
untested
Connect and automate 500+ business appspluginOpsL2
connect-apps · Integrating multi-app orchestration into Claude Code when unified auth + action catalog saves integration time.
untested
Automate pull request creation workflowpluginEngineeringL2
create-pr · Automating PR workflow end-to-end when code style consistency + branch hygiene are mandatory before review.
untested
Analyze and accelerate coding skillspluginProductivityL2
developer-growth-analysis · Personalized developer coaching when automated pattern detection from lived coding behavior beats static curricula.
untested
pr-stamp-sweepworkflow
pr-stamp-sweep · Two-phase fan-out: per-PR reviewer + adversarial security verifier, typed schemas
untested
rue-fix-cycleworkflow
rue-fix-cycle · Real multi-worktree fix harness but tightly bound to Rue compiler repo
untested
release-readinessworkflow
release-readiness · 5-phase release audit with adversarial verify; partly Skein-specific commands
untested
retrospectionworkflow
retrospection · Self-improving harness via RHO; generic, schema-validated, arXiv-backed design
untested
gpu-math-attackworkflow
gpu-math-attack · Mine-then-refute research harness; solid but scoped to one June-22 GPU push
untested
dev-master-orchestratorworkflow
dev-master-orchestrator · Parameterized 5-stage impl→review→verify→amend loop; generic and well-engineered
untested
enrich-companiesworkflow
enrich-companies · Generic multi-dim company enrichment fan-out; schema-validated, model-configurable
untested
examiner-simworkflow
examiner-sim · Real 5-agent parallel rubric grader but tightly coupled to one course repo.
untested
pci-complianceworkflow
pci-compliance · Full PCI SSS v2.0 adversarial assessment; PDF+JSON+CSV out; broadly reusable.
untested
product-critiqueworkflow
product-critique · 9-lens adversarial critique with PR/issue output; scoped to one private app.
untested
sm-ui-buildworkflow
sm-ui-build · Judge-panel UI build with ADR gate; experimental spike, tied to private repo spec.
untested
audit-egressworkflow
audit-egress · Runtime preset discovery + blind adversarial verify; reusable egress audit harness.
untested
grind-baseworkflow
grind-base · Generic multi-round improvement loop with worktree isolation; explicitly a reusable base.
untested
audit-gatesworkflow
audit-gates · Reusable adversarial blast-radius audit: N parallel reviewers, per-finding verifiers, synthesis.
untested
gate-assemblerworkflow
gate-assembler · Real parallel PR gate legs but tightly coupled to wordingone/ember repo specifics.
untested
spark-sweepworkflow
spark-sweep · Sophisticated 9-class scout fan-out with chunking, resume, temporal fence, and JSON contracts.
untested
phase-gateworkflow
phase-gate · Reusable phase-boundary gate: parallel inspectors, adversarial verify, manifest, verdict logic.
untested
product-emitworkflow
product-emit · PR-per-group emitter with model tiering; depends on companion product-critique draft file.
untested
arc-discoveryworkflow
arc-discovery · Structured 3-phase adversarial triage with materiality tripwire schema; reusable decision-gating harness.
untested
plan61-implement-sliceworkflow
plan61-implement-slice · Solid TDD slice workflow but hardcoded to one Rust/Bazel plan in one private repo.
untested
create-ticketsworkflow
create-tickets · 5-phase proposal-to-ticket pipeline with parallel investigation and structured AC schemas; reusable.
untested
design-reviewworkflow
design-review · Route×lens fan-out with adversarial verify phase and prioritized report output; reusable UI harness.
untested
port-cucumber-to-playwrightworkflow
port-cucumber-to-playwright · Well-structured 6-phase port with authority rules, but tied to code.org monorepo conventions.
untested
citation-verifyworkflow
citation-verify · Parallel web-verification harness with strict schema and evidence classification; reusable research sub-step.
untested
rue-overnight-huntworkflow
rue-overnight-hunt · Parameterized bug-hunt fan-out with repro-verified findings; genuinely reusable pattern
untested
fact-grounded-fixworkflow
fact-grounded-fix · Post-mortem-encoded adversarial verify loop; reusable across any codebase for grounded fixes
untested
mlx-movie-director-run-self-improve-imageworkflow
mlx-movie-director-run-self-improve-image · Solid VLM-review loop but tightly coupled to custom run.py; limited reuse outside that repo
page 160 / 161