cyberneticlibrary

The library

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

WORKS 48
Evaluate and improve resumesubagentHRL1
resume-evaluator · strengthening resume impact and passing recruiter scan tests
WORKS 48
Verify phase plans achieve goalssubagentOpsL2
gsd-plan-checker · catching plans that look complete but won't actually achieve phase goals
WORKS 48
Build React components and UI testssubagentEngineeringL2
frontend-builder · building React UI following approved designs and locked API contracts
WORKS 48
Research technology best practicessubagentEngineeringL1
tapps-researcher · researching framework documentation and domain best practices
WORKS 48
Manage project-scoped subagentssubagentOpsL2
README · coordinating parallel frontend, auth, and design work on the same feature
WORKS 48
Implement features test-firstsubagentEngineeringL2
manager-tdd · building new features incrementally with tests proving each step
WORKS 48
Audit application security changessubagentEngineeringL2
security-reviewer · catching multi-tenant isolation bugs and credential-handling vulnerabilities
WORKS 39
Review code against MIAGE criteriasubagentEngineeringL1
code-reviewer-ter · finding correctness bugs and design improvements before code review
WORKS 48
Verify chart visuals across scenariossubagentEngineeringL2
chart-qa-verifier · validating dashboard visualizations match requirements and data integrity
WORKS 48
Audit BountyStrike security defectssubagentEngineeringL2
security-reviewer · auditing security policies for credential handling and secrets rotation
WORKS 39
Execute Opening Range Breakout spreadssubagentFinanceL3
pro-orb · refining product narrative and pitch positioning
WORKS 48
Build decision models and run simulationssubagentOpsProductL3
modeler · capturing and validating system architecture and design decisions
WORKS 48
Develop Dingo language featuressubagentEngineeringL3
golang-developer · implementing Go microservices with idiomatic patterns
WORKS 48
Audit framework reuse opportunitiessubagentEngineeringL2
zipline-researcher · integrating Zipline data pipeline and execution framework
WORKS 48
Classify procedural content generation queriessubagentL1
procgen-classifier · Distinguishing procedural generation questions from general graphics/ML queries before dispatching to specialized PCG agent
WORKS 48
Design real-time event architecturessubagentEngineeringL3
realtime-architect · Architecting presence systems and collaborative editing when latency < 5s and horizontal scaling needed
WORKS 48
Review Python code style and idiomssubagentEngineeringL2
deep-review-python · Anchoring Python style findings in public PEPs (8/20/257) during code review
WORKS 48
Verify RFC and ADR compliancesubagentEngineeringL2
compliance-checker · Catching semantic spec violations—code that references a clause but violates its intent
WORKS 48
Generate JSON Schema definitionssubagentEngineeringL2
fd-json-schema-gen · Catching API contract bugs where generated schema rejects valid v2 payloads or accepts malformed v1
WORKS 53
Diagnose and fix complex bugssubagentEngineeringL3
debugger · Production incidents or complex bugs where root cause is unclear and multi-hypothesis investigation needed
WORKS 48
Resolve review tickets and regressionssubagentEngineeringL3
reviewer · Narrowly scoped second pass after fixes when prior review evidence exists
WORKS 48
Build drug-claim knowledge graphssubagentEngineeringDataL3
kg-agent · Knowledge graph construction from domain-specific structured data when precision is measurable
WORKS 48
Discover and rank data sourcessubagentDataProductL2
explorer · Starting a research project when you need to know what data exists and how hard it is to access
WORKS 48
Conduct peer-reviewed literature surveyssubagentDataProductL3
academic-researcher · Literature assembly when you need verifiable, locally-cached citations with evidence of download success
WORKS 48
Audit dependencies and licensessubagentEngineeringL2
qa-dependency-audit · Pre-deployment QA when you need a single report combining vulnerabilities, outdated packages, and strictness findings
WORKS 50
Review Release Please version bumpssubagentEngineeringL2
release-pr-reviewer · Gate releases when version consistency across files and changelog completeness must be verified
WORKS 48
Build 3D mapping UI with WebSocketssubagentEngineeringProductL3
frontend-3d-builder · Building browser-based 3D visualizations when performance budget is tight and delivery is web-only
WORKS 48
Enforce supply chain dependency rulessubagentEngineeringL2
auditor-supplychain · Periodic supply chain reviews when vendor compliance and risk must be assessed systematically
WORKS 48
Validate agent cell configurationsubagentEngineeringL2
cell-config-auditor · Configuration validation when cells have interdependencies and schema alone is insufficient
WORKS 48
Review alert rule semanticssubagentOpsProductL3
alert-engine-expert · Setting up production monitoring when alerts must minimize noise while catching real incidents
WORKS 48
Design features before implementationsubagentEngineeringProductL2
architect · Major system design when scale, latency, and reliability must be balanced against cost and complexity
WORKS 48
Evaluate technologies and frameworkssubagentEngineeringProductL2
tech-stack-evaluator · Making technology choices when multiple viable stacks exist and risk/learning time must be weighed
WORKS 48
Find and fix stale documentationsubagentEngineeringL2
documentation · Creating maintainable documentation when API stability is high and examples must stay in sync with code
WORKS 48
Run full pre-commit checkssubagentEngineeringL2
pre-commit · Preventing common errors (linting, format, secrets) from entering main branch when consistency is valued
WORKS 48
Create reusable Claude Code skillssubagentEngineeringL3
voyager-skill-creator · Automating multi-step recurring workflows without writing shell scripts directly
WORKS 48
Audit code performance defectssubagentEngineeringL1
performance · Catching O(n²) loops, N+1 queries, and async-blocking patterns that pass tests but fail in production
WORKS 48
Generate new Claude Code subagentsubagentEngineeringL1
zzaia-meta-agent · Creating new specialized agents without manually writing configuration boilerplate
WORKS 48
Validate HTTP test naming conventionssubagentEngineeringL1
spec-naming-validator · Enforcing consistent BDD-style test naming across multi-component C# test suites
WORKS 48
Research domain evaluation criteriasubagentProductL1
gsd-domain-researcher · Surfacing what actual practitioners care about before turning eval criteria into measurable rubrics
WORKS 48
Review industrial construction planssubagentOpsL1
industrial-specialist · Coordinating industrial MEP and material-handling design in large warehouse/manufacturing projects
WORKS 48
Engineer prompts for Flux generationsubagentEngineeringL1
fd-prompt-engineering · Catching ambiguous instructions that will cause agents to behave unexpectedly in production
WORKS 48
Monitor production health and errorssubagentEngineeringL2
site-keeper · Proactive nightly health checks that catch errors before incidents vs. reactive on-call firefighting
WORKS 48
Build hotel invoice backend APIssubagentEngineeringL3
backend-dev · Rapid backend implementation with persistent context to avoid repeating work across sessions
WORKS 48
Audit code for bugs and regressionssubagentEngineeringL1
bug-auditor · Systematic bug hunt across a codebase before major releases
WORKS 48
Migrate Nunjucks to WebC componentssubagentEngineeringL3
eleventy-webc-migrator · Scaling Eleventy sites with modern component scoping and asset bundling
WORKS 48
Build FastAPI backend servicessubagentEngineeringL3
fastapi-backend-expert · Rapid FastAPI backend development with multi-model fallback and user-provided API key management
WORKS 50
Review security and secrets in PRssubagentEngineeringL1
sv-security-reviewer · Pre-merge security gate for open-source projects shipping to app stores
WORKS 48
Generate feature documentationsubagentProductL2
prd-feature · Keeping PRD documentation synchronized with active development
WORKS 48
Observe multi-agent system behaviorsubagentEngineeringL2
heady-observer · Monitoring multi-agent workflows for drift or unexpected behavior changes
WORKS 46
Coordinate multi-agent code reviewssubagentEngineeringL3
swarm-pr · Parallel specialist reviews (security, perf, correctness) in one coordinated workflow
page 4 / 10