cyberneticlibrary

The library

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

ArtifactsCapabilitiesRecipes
kindallskillworkflowmcp_servercommandsubagentplugin|✓ proven✗ harmfulmeasured only198 artifacts
forSalesMarketingHRFinanceLegalOpsProductEngineeringDataProductivitySupportsetup≤ plug & play≤ + a key≤ multi-tool
WORKS 55
Develop from specifications with review loopsworkflowEngineeringProductL4
spec-driven-dev · Driving code generation from formal specs ensuring spec-code correspondence.
WORKS 55
Research topics and synthesize findingsworkflowProductL3
kiln-research · Deep research investigations with parallel multi-perspective evidence gathering.
WORKS 55
Score homepage quality metricsworkflowProductL2
eval-access-fn · Executing complex multi-phase workflows with typed outputs.
WORKS 55
Research topics with web sourcesworkflowProductL2
deep_research · Executing complex multi-phase workflows with typed outputs.
WORKS 58
Cluster feedback into themesworkflowProductOpsL2
feedback-themes · Digesting user feedback batches without manual labeling.
WORKS 55
Generate German learning programworkflowOpsProductL3
german-program · Managing multi-stage product lifecycles with validation gates between phases.
WORKS 54
Run design-to-spec workflowworkflowProductEngineeringL4
az-design-room · Featured lore design and reveal-gate strategy where creative direction + safety must coexist
WORKS 59
Produce validated ebooksworkflowOpsProductL3
ebook · Long-form knowledge products where source verification + adversarial claim-checking beats single-author trust
WORKS 48
Run deterministic loop-until-done verifier workflowworkflowEngineeringProductL3
fierce-ralph · Best for: complex multi-phase orchestration with parallelization and adversarial gates.
WORKS 48
Execute PRISM research and planning workflowworkflowProductEngineeringL3
prototype · Best for: complex multi-phase orchestration with parallelization and adversarial gates.
WORKS 58
Run parallel multi-agent document reviewworkflowEngineeringProductL3
fuyutsui-doc-review · Best for: multi-LLM parallel code review with structural context analysis.
WORKS 57
Audit SPA across dimensions with verificationworkflowEngineeringProductL3
review-spa · Best for: multi-LLM parallel code review with structural context analysis.
WORKS 48
Pipeline PRD through SRS to architectureworkflowProductEngineeringL3
mishkan-init · Best for: complex multi-phase orchestration with parallelization and adversarial gates.
WORKS 54
Perform full iOS mobile audit and redesignworkflowProductEngineeringL3
ios-mobile-audit · Comprehensive systematic audits of code/UI across many files and components.
WORKS 59
Root-cause and design code exploration fixesworkflowEngineeringProductL3
world-class-code-exploration · Building features with test-first discipline across multiple implementation rounds.
WORKS 54
Convert vague goals into testable specsworkflowProductEngineeringL2
goal-lint · Building features with test-first discipline across multiple implementation rounds.
WORKS 60
Run multi-model council debate engineworkflowProductEngineeringL3
council-engine · Multi-model, multi-persona analysis of complex questions with synthesized verdicts.
WORKS 58
Deliver feature end-to-end sequentiallyworkflowProductEngineeringL3
feature-pipeline · Building features with test-first discipline across multiple implementation rounds.
WORKS 55
Run exhaustive multi-agent quality checkworkflowEngineeringProductL3
ultracode-quality-check · Building workflows with structured orchestration and verification gates.
WORKS 55
Research standards and cluster insightsworkflowProductL3
lets-explore · Building workflows with structured orchestration and verification gates.
WORKS 54
Research SPIKE with adversarial verificationworkflowProductL3
spike · Adversarial research workflows that gather, refute, and converge on verified claims.
WORKS 55
Explore pre-commitment questionsworkflowProductL3
explore · Structured agent orchestration with deterministic phase transitions and fault isolation.
WORKS 54
Ground project in verified researchworkflowProductL3
kmaz-research · Adversarial research workflows that gather, refute, and converge on verified claims.
WORKS 58
Deep research with adversarial verificationworkflowProductL3
nihil-odin · When a transformation decision depends on external behavior, upstream docs, or evidence the repo cannot supply.
WORKS 59
Create implementation plans from feature requestsworkflowProductEngineeringL2
implementation_plan · When you want a technical plan before coding, especially for multi-file or unfamiliar code changes.
WORKS 55
Forge and validate ideas across lensesworkflowProductOpsL3
idea-forge · When forging novel research ideas for a specific venue and you want adversarial judges to rank by feasibility and contribution.
WORKS 55
Validate designs through adversarial testingworkflowProductEngineeringL3
adversarial-validation · When you need to verify that tests actually test something and aren't vacuous before committing to green.
WORKS 56
Run fierce debate on high-stakes decisionsworkflowOpsProductL3
fierce-debate · High-stakes architectural or strategic decisions requiring adversarial pressure-testing.
WORKS 59
Evaluate chatbot prompts with consensus panelworkflowProductEngineeringL3
ga-chatbot-qa-panel · Semantic quality assurance using a multi-lens consensus panel.
WORKS 55
Conduct deep research with verified sourcesworkflowProductOpsL3
deep_research · Multi-source research with synthesis and adversarial fact-checking.
WORKS 59
Run weekly site health check reportworkflowOpsProductL3
weekly-checkin · Unattended weekly health checks (GSC, deploy, links, schema, CWV).
WORKS 59
Evaluate neurospatial API design comprehensivelyworkflowProductEngineeringL2
design-review · Multi-phase workflows requiring orchestrated agent handoffs and structured consensus.
WORKS 54
Audit user flows for blockers adversariallyworkflowProductL3
qa-flow-audit · High-stakes architectural or strategic decisions requiring adversarial pressure-testing.
WORKS 54
Design multi-agent tournament evaluationworkflowProductEngineeringL3
tournament-design · Design decisions with multiple valid approaches needing synthesis through pairwise-judged competition.
WORKS 57
Validate spec and plan adversariallyworkflowProductEngineeringL3
superpower-validate · Spec/plan validation before build when requirements are vague or dependencies cross multiple systems.
WORKS 55
Analyze site structure deeplyworkflowProductOpsL3
site-deep-analyze · Deep site analysis combining Playwright crawl, static DOM/CSS inspection, and vision-based UI classification.
WORKS 55
Tune Codex theme iterativelyworkflowEngineeringProductL2
codex-theme-tune · Orchestrating multi-phase autonomous workflows where fan-out and synthesis improve outcomes.
WORKS 57
Tune Antigravity theme iterativelyworkflowEngineeringProductL2
antigravity-theme-tune · Orchestrating multi-phase autonomous workflows where fan-out and synthesis improve outcomes.
WORKS 55
Triage feature scope and designworkflowProductEngineeringL3
fathomdb-v05-feature-triage · Orchestrating multi-phase autonomous workflows where fan-out and synthesis improve outcomes.
WORKS 60
Plan phased build for collaborative editorworkflowProductL2
notion-clone-build-plan · Workflow notion clone build plan with multi-agent verification.
WORKS 58
Redesign UI with Tailwind and shadcnworkflowProductL2
ui-redesign · Workflow ui redesign with multi-agent verification.
WORKS 60
Migrate UI primitives to Lumina kitworkflowProductL3
migrate-engineering-primitives · Workflow migrate engineering primitives with multi-agent verification.
WORKS 59
Generate ideas, dedupe, filter by rubricworkflowProductL2
generate-and-filter · Brainstorming where you want breadth first, then only the highest-quality, non-duplicate, rubric-passing ideas. Pass a bare topic string, or {topic, generators?, rubric?, threshold?, keep?} as args.
WORKS 54
Audit feature gaps versus roadmapworkflowProductL3
feature-gap-review · Multi-dimensional acceptance criteria requiring independent per-AC verification.
WORKS 53
Sweep GUI surfaces for jank and issuesworkflowProductL2
gui-audit · Triage workflows where items need independent lens-based analysis before synthesis.
WORKS 59
Implement remaining PRD slicesworkflowProductL3
ship-prd · Topologically sorted task execution across human-in-the-loop and agentic work layers.
WORKS 58
Implement and verify pending milestonesworkflowProductEngineeringL3
argus-implement · Complex workflows requiring parallel agents, synthesized judgment, or multi-phase triage.
WORKS 55
Improve game dictionaries with dual-judge validationworkflowProductEngineeringL3
dictionary-improvement · Complex workflows requiring parallel agents, synthesized judgment, or multi-phase triage.
WORKS 55
Audit active and parked epicsworkflowProductEngineeringL3
epic-audit · Triage workflows where items need independent lens-based analysis before synthesis.
WORKS 55
Validate product demand with sourced evidenceworkflowProductMarketingL3
market-validation-research · Gathering and adversarially verifying market or technical evidence from multiple sources with cited claims.
page 1 / 4→