Skill Catalogue

Skill Catalogue

Complete list of canonical skills in the registry.

Important

The canonical source of truth is always the modular node files under registry/nodes/**/*.json, consolidated into registry/gaia.json. If this page disagrees with the nodes, the nodes win. All IDs use kebab-case. Under the G7 framework, generic skills are starless (rank-less) and represent abstract capabilities; only named contributor implementations have star ratings (levels).

---

Basics

Primitive building blocks. These are the fundamental capabilities with no prerequisites in the graph.

ID	Name	Description
adaptive-pattern-learning	Adaptive Pattern Learning	Records task outcomes and identifies recurring scenarios to rank optimal response strategies over time. It enables genuine self-improvement through a feedback learning loop by matching scenarios and transferring knowledge across related domains via vector similarity.
api-call	API Call	Constructs, executes, and handles responses from HTTP APIs by interpreting documentation and selecting appropriate endpoints and parameters. This skill allows agents to fetch dynamic data, trigger remote actions, and interact with external systems programmatically by adhering to OpenAPI or other API specifications.
audience-model	Audience Model	Adapts tone, complexity, and framing of output to match a target audience profile.
autonomous-engineering-platform	Autonomous Engineering Platform	An autonomous engineering platform integrating chat, specs, tasks, and code for full agentic software delivery.
bioinformatic-sequence-analysis	Bioinformatic Sequence Analysis	Performs biological sequence alignment, similarity searches, multiple sequence alignment (MSA), and genomic variant analysis using standard bioinformatics tools and databases.
browser-control	Browser Control	Directly controls a web browser via low-level protocols (like CDP) to manipulate the DOM, manage cookies, intercept network traffic, and simulate user interactions (clicking via coordinates, taking screenshots) without relying on heavy wrapper frameworks like Playwright or Puppeteer.
chain-of-thought	Chain-of-Thought Reasoning	Produces explicit intermediate reasoning steps before arriving at a final answer, dramatically improving accuracy on multi-step problems.
chunk-document	Chunk Document	Splits a document into semantically meaningful segments optimized for embedding and retrieval.
cite-sources	Cite Sources	Attributes claims to specific sources with proper references, URLs, or bibliographic entries.
classify	Classify	Assigns one or more categorical labels to an input based on learned or rule-based criteria.
cli-modernization	CLI Modernization	Refactors command-line interfaces for improved UX, plugin architecture, and extensibility.
clinical-data-retrieval	Clinical Data Retrieval	Retrieves and parses clinical trial records, drug approvals, regulatory filings, or medical variant databases (e.g. ClinicalTrials.gov, OpenFDA, ClinVar).
code-execution	Code Execution	Writes and executes code in a sandboxed environment, uses the runtime output to verify correctness, and iterates until the result is correct.
code-explain	Code Explain	Generates accurate natural-language explanations of source code, describing intent, logic flow, and key decisions at function or module level.
code-generation	Code Generation	Produces syntactically correct and functionally appropriate source code from specifications or prompts.
computer-use	Computer Use	Controls desktop GUIs and web browsers by interpreting screenshots, issuing mouse/keyboard actions, and verifying visual state to complete open-ended computer tasks.
context-compression	Context Compression	Reduces the length of prompts or retrieved context to fit token limits while preserving semantic content, using techniques such as selective token removal, summarization, or token-classification-based pruning (e.g. LLMLingua).
core-platform-implementation	Core Platform Implementation	Implements foundational platform architecture including plugin discovery, server lifecycle management, and API contracts.
cultural-localization	Cultural Localization	Adapts content for regional audiences beyond translation: idioms, cultural nuance, tone, trust conventions, and regional compliance for culturally faithful output.
data-visualize	Data Visualize	Generates charts, graphs, and visual summaries from datasets by selecting appropriate visualization types and mapping data dimensions.
diff-content	Diff Content	Compares two versions of content and produces a structured delta highlighting additions, deletions, and modifications.
document-editing	Document Editing	Reads, edits, repacks, and applies styling or design principles to structured binary document formats such as PPTX, DOCX, and XLSX.
domain-modeling	Domain Modeling	Build and maintain a shared domain model with ubiquitous language and architectural decision records.
embed-text	Embed Text	Converts text into dense vector representations suitable for similarity search and clustering.
error-interpretation	Error Interpretation	Diagnoses error messages, stack traces, and failure modes to identify root causes and suggest fixes.
evaluate-output	Evaluate Output	Assesses the quality, correctness, or safety of generated output against defined criteria.
extract-entities	Extract Entities	Identifies and extracts named entities such as people, organizations, dates, and locations from text.
feed-monitoring	Feed Monitoring	Monitors RSS, Atom, blog, or other recurring content feeds, discovers updates, tracks read state, and surfaces new signals for downstream agent workflows.
few-shot-learning	Few-Shot Learning	Conditions a language model on a small number of input-output demonstrations within the prompt, enabling task adaptation without weight updates.
fine-tune	Fine-Tune	Adapts a pre-trained model to a new task or domain by updating model weights using parameter-efficient methods such as LoRA, QLoRA, or full supervised fine-tuning, without retraining from scratch.
format-output	Format Output	Structures raw output into a specified format such as markdown, JSON, CSV, or HTML.
framework-upgrade	Framework Upgrade	Guides an agent through upgrading a project from one major framework version to another, including breaking-change detection, migration steps, and post-upgrade validation.
generate-sql	Generate SQL	Translates natural-language data questions into syntactically correct, executable SQL queries against a given schema.
generate-test	Generate Test	Analyzes source code or natural-language specifications to identify functions, classes, and their dependencies, and generates thorough, maintainable test suites. It handles test case structuring, setup and teardown fixtures, mocking external dependencies, and validating normal operation alongside boundary and error conditions.
generate-text	Generate Text	Produces coherent natural language output given a prompt, instruction, or context window.
genomic-data-retrieval	Genomic Data Retrieval	Queries and retrieves DNA sequence data, transcription factor binding models, cis-regulatory elements, and genetic annotations from genomic repositories (e.g. Ensembl, GTEx, JASPAR, UniBind, ENCODE, UCSC, dbSNP).
git-diff-risk-analysis	Git Diff Risk Analysis	Analyzes git diffs for complexity, churn, and risk scores to prioritize review attention and flag dangerous changes.
headless-worker-collect	Headless Worker Collect	Aggregates and formats results from distributed headless workers stored in shared memory with filtering and status reporting.
headless-worker-spawn	Headless Worker Spawn	Launches headless AI worker processes for parallel background task execution with configurable worker types and shared memory coordination.
hybrid-workflow-coordination	Hybrid Workflow Coordination	Routes tasks between interactive reasoning phases and parallel background execution with configurable workflow templates.
hypothesis-generate	Hypothesis Generation	Formulates novel, testable scientific hypotheses by synthesising existing literature, identifying knowledge gaps, and proposing mechanistic explanations.
image-caption	Image Caption	Generates accurate natural-language descriptions of images, capturing objects, actions, and spatial relationships.
image-generate	Image Generate	Creates photorealistic or stylized images from text prompts using diffusion-based or autoregressive generative models.
issue-triage	Issue Triage	Classifies incoming issue reports through a structured state machine, assigns triage roles (bug/enhancement, needs-info/ready-for-agent/wontfix), reproduces bugs, requests missing detail, and produces structured resolution briefs for agent or human handoff.
literature-search	Literature Search	Queries and searches academic literature databases (e.g. PubMed, arXiv, bioRxiv, OpenAlex) to locate papers, retrieve abstracts, resolve citations, and fetch full-text documents.
logical-inference	Logical Inference	Applies deductive, inductive, or abductive reasoning to derive valid conclusions from premises and structured knowledge.
math-reason	Math Reason	Solves multi-step mathematical problems including arithmetic, algebra, calculus, and competition mathematics through symbolic and numeric reasoning.
mcp-integration	MCP Integration	Connect to and invoke tools exposed by Model Context Protocol (MCP) servers — enumerate available tools, execute calls, and handle responses across any MCP-compatible backend.
memory-pattern-design	Memory Pattern Design	Designs recurring memory storage patterns for AI agents with LRU caching, SQLite persistence, and associative retrieval.
molecular-databases	Molecular Databases	Queries and retrieves structural, chemical, and biological activity data for small molecules, chemical compounds, or drug candidates from molecular repositories (e.g., ChEMBL, PubChem).
multi-repo-coordination	Multi-Repo Coordination	Manages synchronized operations across multiple repositories including cross-repo PRs, dependency tracking, and bulk workflow automation.
object-detection	Object Detection	Locates and classifies multiple objects within images by producing bounding boxes, confidence scores, and category labels in a single forward pass.
optical-character-recognition	OCR	Extracts machine-readable text from raster images, scanned pages, and photo documents using optical character recognition, preserving layout and handling skew, noise, and varied fonts.
parallel-execution	Parallel Execution	Decompose a task into independent sub-tasks and execute them concurrently, merging results with configurable concurrency limits and queue-based state tracking.
parse-html	Parse HTML	Extracts structured content from raw HTML, navigating DOM trees and handling malformed markup.
parse-json	Parse JSON	Extracts structured data from JSON-formatted input, handling nested objects and arrays.
parse-pdf	Parse PDF	Extracts text, tables, equations, and structure from PDF documents, preserving layout and reading order.
pathway-ontology-retrieval	Pathway Ontology Retrieval	Queries, parses, and resolves biological pathway models, cellular reactions, and hierarchical ontology metadata from pathway and ontology search systems (e.g., Reactome, QuickGO, EMBL-EBI OLS).
performance-tuning	Performance Tuning	Profiles execution hotspots, measures throughput and latency, and applies targeted optimizations to reduce resource consumption.
plan-decompose	Plan and Decompose	Breaks a complex objective into an ordered sequence of executable sub-tasks.
probabilistic-programming	Probabilistic Programming	Specifies Bayesian models and runs inference (MCMC/NUTS, variational) to quantify epistemic and aleatoric uncertainty using PyMC, Stan, Pyro, or NumPyro.
prompt-injection-defense	Prompt Injection Defense	Detects and neutralizes adversarial instructions injected into agent context from untrusted external sources (indirect prompt injection), using techniques such as context isolation, hierarchical intent verification, or semantic consistency checks.
protein-structure-analysis	Protein Structure Analysis	Analyzes, searches, and aligns three-dimensional macromolecular structures using structural databases, alignment search algorithms (e.g. Foldseek, PDB search), or molecular modeling tools.
proteomic-data-retrieval	Proteomic Data Retrieval	Queries and retrieves amino acid sequences, protein families, functional domains, tissue distribution, and protein-protein interactions from proteomic repositories (e.g. UniProt, InterPro, Human Protein Atlas, STRING).
question-answer	Question Answer	Produces accurate, context-grounded answers to natural-language questions, handling unanswerable cases.
rank	Rank	Orders a set of candidate items by relevance, quality, or fitness for a given objective.
refactor-code	Refactor Code	Restructures existing source code to improve readability, maintainability, or performance without changing observable behavior.
requirements-analysis	Requirements Analysis	Elicit and structure requirements from stakeholder inputs into formal specifications — user stories, acceptance criteria, and traceability matrices.
retrieve	Retrieve	Fetches relevant documents or passages from an indexed corpus given a query.
reward-modeling	Reward Modeling	Learns a scalar reward signal from human preference comparisons between model outputs, enabling reinforcement learning from human feedback (RLHF) to align model behavior with human values.
route-intent	Route Intent	Classifies user intent and directs execution to the appropriate handler, tool, or sub-agent.
schema-design	Schema Design	Design database schemas across relational, NoSQL, graph, and time-series stores — entity modelling, normalization, indexing strategies, and migration planning.
scientific-visualization	Scientific Visualization	Creates publication-ready scientific figures with multi-panel layouts, statistical annotations, colorblind-safe palettes, and journal-specific formatting (Nature, Science, Cell) using matplotlib, seaborn, and plotly. It also includes visualizing, analyzing, and rendering protein and molecular structures without a hardware-accelerated GUI.
score-relevance	Score Relevance	Assigns a numerical relevance score to candidate items relative to a query or objective.
self-consistency	Self-Consistency	Samples multiple independent reasoning paths for the same problem and selects the answer by majority vote, improving robustness without any additional training.
self-critique	Self-Critique	Iteratively evaluates and refines its own outputs using self-generated feedback, improving quality without external supervision.
semantic-cache	Semantic Cache	Stores LLM responses keyed by embedding similarity so that semantically equivalent queries are served from cache, reducing inference latency and token cost without sacrificing answer quality.
sentiment-analysis	Sentiment Analysis	Classifies the affective polarity (positive / negative / neutral, or fine-grained) of user-generated text. Covers pipelines from raw noisy input through preprocessing, inference (Transformer, Lexicon, or LLM), and output normalisation.
sequential-agent-pipeline	Sequential Agent Pipeline	Chains agent outputs sequentially so each step's result feeds the next, enabling multi-stage data transformation and coordinated processing.
skill-discovery	Skill Discovery	Searches a skill or tool registry, ranks results by relevance and install count, and surfaces candidates for the agent to adopt or invoke.
software-design	Software Design	Design software modules with clean interfaces, seams, and deep implementations for maintainability and testability.
speech-to-text	Speech to Text	Transcribes spoken audio into accurate text, handling diverse accents, noise conditions, and multiple languages.
statistical-analysis	Statistical Analysis	Performs hypothesis testing, regression analysis, and Bayesian modelling with effect size calculation and APA-formatted statistical reporting using scipy, statsmodels, and PyMC.
structured-output	Structured Output Generation	Generates output guaranteed to conform to a given schema (JSON, YAML, Pydantic model, etc.) using constrained decoding or grammar-guided generation.
summarize	Summarize	Condenses longer input into a shorter representation that preserves key information and intent.
swarm-topology-management	Swarm Topology Management	Initializes and manages multi-agent swarm network topologies (hierarchical, mesh, ring, star) with load balancing and fault tolerance.
synthetic-data-generation	Synthetic Data Generation	Generates privacy-preserving, domain-representative datasets via tabular synthesis and generative models, infusing schema and regulatory constraints to substitute for sensitive real data.
system-integration	System Integration	Connects disparate subsystems via shared contracts, event buses, and compatibility layers for coherent cross-component operation.
test-driven-development	Test-Driven Development	Enforces a strict red-green-refactor TDD loop: writes failing tests before implementation, blocks code generation that skips the test step, and validates coverage thresholds before marking tasks complete.
text-to-speech	Text to Speech	Synthesizes natural-sounding speech audio from text input, supporting voice cloning and prosody control.
time-series-forecasting	Time Series Forecasting	Forecasts temporal sequences with seasonal-trend decomposition, ensemble models (ARIMA, Prophet, NBEATS, TFT), and residual-based anomaly detection.
token-observability	Token Observability	Tracks and analyzes token usage and cost across multiple AI models and agents, providing session-level and project-level spending metrics. Operates locally by reading session data directly from disk and utilizes local pricing calculation, ensuring no proxies or API keys are required. Features CLI commands to optimize token waste, compare models, and track yield.
tokenize	Tokenize	Splits input text into discrete tokens suitable for downstream processing by language models.
tool-select	Tool Select	Chooses the most appropriate tool or API from a set of available options given a task description.
tool-use	Tool Use	Invokes external functions or APIs by generating well-formed call signatures, parsing results, and incorporating them into reasoning.
translate	Translate	Converts text from one natural language to another while preserving meaning, tone, and formatting.
ux-audit	UX Audit	Systematically evaluates a user interface against established usability heuristics or accessibility standards, producing a scored finding report with remediation recommendations.
vector-db-optimization	Vector DB Optimization	Tunes vector database indices, sharding strategies, and multi-database coordination for production-scale retrieval.
vector-search	Vector Search	Performs similarity search over high-dimensional embeddings using cosine, Euclidean, dot-product, or custom distance metrics.
vision-qa	Visual Question Answering	Answers natural-language questions grounded in a specific image or screenshot by combining visual perception with language reasoning, enabling visual debugging, UI analysis, and document-image Q&A.
web-search	Web Search	Queries external search engines or APIs and retrieves relevant result sets.
worker-agent-dispatch	Worker Agent Dispatch	Maps trigger events to optimal agent combinations for background task execution with performance tracking and continuous feedback.
write-report	Write Report	Produces structured, multi-section written output with headings, citations, and coherent narrative.

---

Extras

Skills that emerge from the combination of two or more prerequisite skills.

ID	Name	Prerequisites
advanced-swarm-coordination	Advanced Swarm Coordination	swarm-topology-management · multi-agent-debate
adversarial-robustness-testing	Adversarial Robustness Testing	agent-eval · guardrails · prompt-injection-defense
agent-environment-setup	agent-environment-setup	document-editing · tool-use
agent-eval	Agent Evaluation	evaluate-output · score-relevance
agent-handoff	agent-handoff	summarize · document-editing · context-compression
agent-memory-learning	Agent Memory Learning	adaptive-pattern-learning · memory-manage
agent-memory-platform	Agent Memory Platform	distributed-vector-memory · agent-memory-learning · memory-pattern-design · vector-db-optimization · vector-search
agentic-workflow-design	Agentic Workflow Design	workflow-automation · plan-decompose
architecture-diagram	Architecture Diagram	data-visualize · format-output · write-report
automated-testing	Automated Testing	generate-test · code-execution · error-interpretation
autonomous-data-scientist	Autonomous Data Scientist	data-analysis · math-reason · research
autonomous-debug	Autonomous Debug	code-generation · code-execution · error-interpretation
autonomous-web-research	Autonomous Web Research	ghostwrite · knowledge-harvest · research · web-scrape · web-search
brainstorming	Brainstorming	requirements-analysis · question-answer
browser-automation	Browser Automation	web-search · computer-use
career-operations	Career Operations	autonomous-web-research · summarize · document-editing
causal-inference	Causal Inference	hypothesis-generate · logical-inference · statistical-analysis
cloud-platform-management	Cloud Platform Management	api-call · tool-use
code-review-pipeline	Code Review Pipeline	code-generation · diff-content · evaluate-output
collaborative-diagramming	Collaborative Diagramming	api-call · data-visualize
computational-biology-workflows	Computational Biology Workflows	bioinformatic-sequence-analysis · genomic-data-retrieval · molecular-databases · protein-structure-analysis · proteomic-data-retrieval
content-moderation	Content Moderation	classify · sentiment-analysis · extract-entities
conversational-agent	Conversational Agent	question-answer · memory-manage · route-intent
data-analysis	Data Analysis	generate-sql · data-visualize · summarize
deployment-automation	Deployment Automation	workflow-automation · code-execution
design-generation	Design Generation	code-generation · vision-qa
design-review	Design Review	evaluate-output · plan-decompose
design-system-extraction	Design System Extraction	browser-automation · parse-html
detect-anomaly	Detect Anomaly	evaluate-output · error-interpretation
dispatching-parallel-agents	Dispatching Parallel Agents	plan-decompose · parallel-execution
distributed-consensus-coordination	Hive Mind Coordination	swarm-topology-management · worker-agent-dispatch
distributed-neural-training	Distributed Neural Training	parallel-execution · embed-text
distributed-vector-memory	Distributed Vector Memory	vector-search · memory-manage
document-analyst	Document Analyst	extract-entities · format-output · optical-character-recognition · parse-json · parse-pdf · summarize
dual-mode	Dual Mode	headless-worker-spawn · headless-worker-collect · hybrid-workflow-coordination
e2e-testing	End-to-End Testing	browser-automation · automated-testing
edge-optimization	Edge Optimization	evaluate-output · fine-tune · performance-tuning
engineering-discipline	mattpocock-engineering	autonomous-debug · refactor-code · vertical-slice-planning · prd-generation · issue-triage · ubiquitous-language · code-explain · grill-with-docs · prototype · agent-environment-setup
executing-plans	Executing Plans	writing-plans · tool-use
explainability-audit	Explainability Audit	data-visualize · evaluate-output · statistical-analysis
financial-modeling	Financial Modeling	data-analysis · generate-sql · statistical-analysis
finishing-a-development-branch	Finishing a Development Branch	code-review-pipeline · diff-content
full-stack-developer	Full-Stack Developer	code-review-pipeline · automated-testing · refactor-code
function-calling	Function Calling	structured-output · api-call · tool-select
generative-media	Generative Media	api-call · data-visualize
ghostwrite	Ghostwrite	research · write-report · audience-model
git-integration	GitHub Platform Mastery	code-review-pipeline · multi-repo-coordination · deployment-automation · project-management · workflow-automation
graph-driven-issue-triage	Graph-Driven Issue Triage	issue-triage · knowledge-graph-build
grill-me	Grill Me	design-review · self-critique
grill-with-docs	Grill With Docs	grill-me · ubiquitous-language · document-editing
grounding	Grounding	retrieve · cite-sources · evaluate-output
guardrails	Guardrails	evaluate-output · classify · structured-output
humanize-prose	Humanize Prose	document-editing · audience-model · format-output
knowledge-graph-build	Knowledge Graph Construction	extract-entities · logical-inference
knowledge-harvest	Knowledge Harvest	web-scrape · extract-entities · embed-text
knowledge-management	Knowledge Management	api-call · retrieve
literature-review	Literature Review	cite-sources · literature-search · research · summarize
mathematical-animation	Mathematical Animation	code-execution · data-visualize
mcp-debugger-control	MCP Debugger Control	mcp-integration · refactor-code
mcp-server-creation	MCP Server Creation	mcp-integration · tool-creation · api-call
memory-manage	Memory Manage	context-compression · retrieve
ml-artifact-management	ML Artifact Management	api-call · classify
ml-pipeline	ML Pipeline	data-analysis · automated-testing · code-generation
multi-agent-debate	Multi-Agent Debate	self-critique · evaluate-output · chain-of-thought
multi-agent-orchestration-v	Multi-Agent Orchestration	plan-and-execute · route-intent · tool-select
multi-node-orchestration	Flow Nexus Orchestration	multi-agent-orchestration-v · cloud-platform-management · distributed-neural-training
multimodal-reasoning	Multimodal Reasoning	image-caption · question-answer · logical-inference
ontology-alignment	Ontology Alignment	extract-entities · knowledge-graph-build · schema-design
personal	mattpocock-personal	document-editing · personal-knowledge-management
personal-knowledge-management	personal-knowledge-management	retrieve · context-compression
plan-and-execute	Plan and Execute	route-intent · plan-decompose · tool-select
platform-modernization-sprint	Platform Modernization Sprint	cli-modernization · core-platform-implementation · system-integration · performance-tuning · advanced-swarm-coordination
prd-generation	PRD Generation	write-report · plan-decompose
prediction-market-analysis	Prediction Market Analysis	data-analysis · web-search · statistical-analysis
productivity	mattpocock-productivity	agent-handoff · context-compression · grill-me · tool-creation
project-management	Project Management	api-call · issue-triage
prompt-optimization	Prompt Optimization	evaluate-output · generate-text
prototype	prototype	code-generation · code-execution
rag-pipeline	RAG Pipeline	retrieve · chunk-document · embed-text · score-relevance · tokenize · rank
re-act-reasoning	ReAct Reasoning	plan-decompose · tool-use
real-time-voice-assistant	Real-Time Voice Assistant	voice-agent · memory-manage · plan-and-execute
reasoning-pattern-bank	Reasoning Pattern Bank	adaptive-pattern-learning · agent-memory-learning
receiving-code-review	Receiving Code Review	requesting-code-review · verification-before-completion
recursive-self-improvement	Recursive Self-Improvement	autonomous-debug · evaluate-output · plan-and-execute
registry-curation	Registry Curation	research · code-generation · code-execution
registry-entry-audit	Gaia Audit	retrieve · cite-sources · evaluate-output
registry-health-scan	Gaia Meta Audit	registry-entry-audit · registry-curation · detect-anomaly
regulatory-compliance-mapping	Regulatory Compliance Mapping	extract-entities · knowledge-graph-build · security-audit
release-automation	Release Automation	workflow-automation · code-execution · generate-text
requesting-code-review	Requesting Code Review	code-review-pipeline · dispatching-parallel-agents
research	Research	web-search · summarize · cite-sources
scientific-discovery	Autonomous Scientific Discovery	computational-biology-workflows · hypothesis-generate · math-reason · research
scientific-writing	Scientific Writing	write-report · cite-sources · scientific-visualization
search-engine-optimization	search-engine-optimization	parse-html · parse-json
security-audit	Security Audit	code-review-pipeline · evaluate-output
skill-authoring	Skill Authoring	generate-text · structured-output · code-generation
skill-performance-benchmarking	Skill Performance Benchmarking	agent-eval · skill-discovery · statistical-analysis
skill-security-analysis	Skill Security Analysis	security-audit · prompt-injection-defense · skill-discovery
stealth-browser-interaction	Stealth Browser Interaction	browser-automation · computer-use
subagent-driven-development	Subagent-Driven Development	executing-plans · dispatching-parallel-agents · code-review-pipeline
supply-chain-optimization	Supply Chain Optimization	data-analysis · detect-anomaly · statistical-analysis
systematic-debugging	Systematic Debugging	hypothesis-generate · error-interpretation
text-to-sql-pipeline	Text-to-SQL Pipeline	generate-sql · parse-json · format-output
threat-intelligence-synthesis	Threat Intelligence Synthesis	feed-monitoring · knowledge-graph-build · security-audit
tool-chaining	Tool Chaining	tool-use · tool-select
tool-creation	Tool Creation	code-generation · tool-use
translation-pipeline	Translation Pipeline	translate · sentiment-analysis · audience-model
tree-of-thought	Tree of Thought	chain-of-thought · plan-decompose
ubiquitous-language	Ubiquitous Language	extract-entities · requirements-analysis
using-git-worktrees	Using Git Worktrees	code-execution · tool-use
verification-before-completion	Verification Before Completion	code-execution · self-critique
vertical-slice-planning	Vertical Slice Planning	plan-decompose · route-intent
video-intelligence	Video Intelligence	multimodal-reasoning · speech-to-text
voice-agent	Voice Agent	speech-to-text · question-answer · text-to-speech
web-accessibility	web-accessibility	browser-control · ux-audit
web-best-practices	web-best-practices	code-review-pipeline · performance-tuning
web-quality-audit	web-quality-audit	registry-health-scan · performance-tuning · search-engine-optimization · web-accessibility · web-best-practices
web-scrape	Web Scrape	web-search · parse-html · extract-entities
wiki-search	Wiki Search	retrieve · embed-text · summarize
workflow-automation	Workflow Automation	plan-decompose · tool-use · api-call
workspace-automation	Workspace Automation	api-call · document-editing
writing-plans	Writing Plans	plan-decompose · requirements-analysis
x-twitter-automation	X/Twitter Automation	browser-automation · web-scrape · workflow-automation

---

Uniques

Specialized basic capabilities that have achieved significant trust magnitudes and are graph-isolated (having zero prerequisites and zero derivatives referenced by other nodes).

There are currently no unique generic skill nodes in the registry.

---

Ultimates

The apex skills of the graph. These complex capabilities bundle dozens of component skills, require strict multi-source evidence, and require verifier sign-offs to claim.

ID	Name	Prerequisites
founder-mode-orchestration	Founder Mode	question-answer · evaluate-output · code-review-pipeline · ux-audit · vertical-slice-planning · design-system-extraction · design-review · code-generation · systematic-debugging · e2e-testing · finishing-a-development-branch · deployment-automation · detect-anomaly · document-editing · browser-automation · security-audit · browser-control · web-scrape · skill-authoring · write-report · multi-agent-debate · skill-performance-benchmarking · automated-testing · mcp-integration · memory-manage · guardrails · knowledge-management · format-output · workspace-automation · project-management · prompt-optimization
multi-topology-orchestration	Ruflo	multi-node-orchestration · agent-memory-platform · git-integration · reasoning-pattern-bank · platform-modernization-sprint · dual-mode
skill-mastery	mattpocock-skills	engineering-discipline · productivity · personal · tool-creation · ubiquitous-language
superpowers	Superpowers	brainstorming · dispatching-parallel-agents · executing-plans · finishing-a-development-branch · receiving-code-review · requesting-code-review · subagent-driven-development · systematic-debugging · using-git-worktrees · verification-before-completion · writing-plans