agentsclimarketplace

Skill connection map

Skill onfire7777/universal-ai-skills-library/skills/skill-connection-map

Maps the hidden connections between all 630+ installed skills, enabling cross-pollination of ideas across domains. Use when you want to find unexpected connections, when a problem in one domain might benefit from thinking in another domain, when you want to transfer patterns between fields, or when you want to see the bigger picture of how skills relate. This is the "neural network" that connects all skills into a living, interconnected knowledge system.From its SKILL.md

Install
npx -y skills add onfire7777/universal-ai-skills-library --skill skill-connection-map

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 14 stars14 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its file declares

Copied from the file, not written here

The file declares its own license as Unspecified. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

12.1 KB, ~2.6k tokens by cl100k_base, as published. Nobody here has run it

Skill Connection Map

Purpose

Individual skills are powerful. But the real magic happens at the intersections — when a security insight improves a design decision, when a creativity technique solves a debugging problem, when a planning framework reveals an architecture flaw. This skill maps those connections and teaches you to find new ones.

"The most exciting phrase in science is not 'Eureka!' but 'That's funny...'" — Isaac Asimov

The best ideas come from connecting things that don't obviously belong together.


The Connection Taxonomy

Skills connect across seven types of relationships. When exploring connections, systematically check each type.

Type 1: Structural Analogy

Two skills from different domains share the same underlying structure or pattern.

Skill ASkill BShared Structure
thinking-systems (feedback loops)simota-bolt (caching invalidation)Both are about managing circular dependencies in complex systems
thinking-first-principles (decompose to fundamentals)pm-epic-breakdown (decompose epics to stories)Both use recursive decomposition to make the complex manageable
thinking-red-team (adversarial thinking)owasp-security (threat modeling)Both systematically think like an attacker to find weaknesses
color-palette (harmony from constraints)architecture-patterns (elegance from constraints)Both create beauty/quality by working within deliberate constraints
wondelai-refactoring-ui (visual hierarchy)sql-optimization (query execution plan)Both are about ordering operations for maximum efficiency of processing
lyn-bayesian-calibration (update beliefs with evidence)test-driven-development (update code with failing tests)Both use a tight feedback loop of hypothesis → evidence → correction

How to use: When stuck in one domain, ask "What skill from a completely different domain has the same underlying structure?" Then apply that skill's techniques with domain-appropriate translation.

Type 2: Inverse/Complement

Two skills that are mirror images — one creates, the other validates; one builds, the other breaks.

Creator SkillValidator SkillRelationship
frontend-designux-auditDesign creates, audit validates
architecture-patternsjeffallan-the-foolArchitecture proposes, The Fool stress-tests
pm-prd-developmentthinking-pre-mortemPRD plans for success, pre-mortem plans for failure
creative-connections-enginemeta-thinking-engineCreativity generates, metacognition evaluates
react-best-practicescode-review-excellenceImplementation follows patterns, review checks adherence
strategic-planning-enginelyn-project-risk-registerPlanning charts the course, risk register maps the dangers
auth-patternstob-codeql-analysisAuth implements security, CodeQL finds the holes

How to use: Every creative/constructive skill should be paired with its inverse. Never build without validating. Never plan without stress-testing. This is the fundamental rhythm of quality work.

Type 3: Prerequisite Chain

Skill A must be applied before Skill B can be effective.

FirstThenWhy
pm-problem-statementpm-prd-developmentYou can't write requirements for a problem you haven't defined
meta-thinking-enginethinking-model-routerUnderstand your thinking mode before selecting a mental model
ui-ux-pro-maxfrontend-designDesign the experience before implementing the interface
color-palettedesign-system-creationGenerate your palette before building the system that uses it
owasp-securitysemgrep-code-securityUnderstand the threats before scanning for them
pm-discovery-processpm-roadmap-planningValidate the problem before planning the solution
systematic-debugginginteractive-debuggerForm a hypothesis before setting breakpoints
thinking-first-principlesthinking-second-orderDecompose to fundamentals, then trace the consequences

How to use: Before applying any skill, ask "Is there a prerequisite skill that should fire first?" Working out of order wastes effort and produces inferior results.

Type 4: Cross-Domain Transfer

A technique from one domain that directly applies in another, often with surprising effectiveness.

Source SkillTarget DomainTransfer
thinking-scientific-method (hypothesis → experiment → conclusion)DebuggingEvery bug fix should be a hypothesis tested against evidence, not a random change
wondelai-never-split-difference (negotiation)Code Review"Mirroring" and "labeling" techniques make code review feedback more effective
thinking-inversion (what would make this fail?)UI DesignDesign for the error state first, then the happy path
thinking-theory-of-constraints (find the bottleneck)Performance OptimizationDon't optimize everything — find and fix the single biggest bottleneck
wondelai-design-everyday-things (affordances)API DesignAPIs should have "affordances" — their interface should suggest correct usage
thinking-ooda-loop (Observe-Orient-Decide-Act)Incident ResponseMilitary decision cycle maps perfectly to production incident management
thinking-margin-of-safety (engineering buffers)Sprint PlanningAlways plan with 20-30% buffer — the margin of safety principle
graphic-design (Gestalt principles)Information ArchitectureProximity, similarity, and closure apply to data organization, not just visual layout
wondelai-hooked (habit loops)Developer ExperienceMake your tools habit-forming: trigger → action → reward → investment
thinking-cynefin (complexity domains)Project ManagementDifferent project types need different management approaches — don't use agile for everything

How to use: This is the most powerful connection type. Actively ask "What would an expert from a completely different field do with this problem?" Then find the skill from that field and apply its principles.

Type 5: Emergent Combination

Two or more skills that, when combined, create a capability that neither has alone.

Skills CombinedEmergent Capability
creative-connections-engine + thinking-model-router + meta-thinking-engineAdaptive Creative Reasoning — Generate creative ideas, route them through the right analytical framework, then metacognitively evaluate the quality of both the ideas and the reasoning
ui-ux-pro-max + color-palette + wondelai-web-typography + interaction-designComplete Design Language — Not just a UI, but a coherent visual language with color theory, typographic rigor, and interaction patterns that feel inevitable
owasp-security + privacy-by-design + auth-patterns + encryption-aesSecurity-First Architecture — Not security bolted on, but security as the foundation that shapes every architectural decision
pm-discovery-process + pm-opportunity-tree + thinking-jobs-to-be-done + wondelai-mom-testEvidence-Based Product Discovery — Discover real problems, map the solution space, validate with users, all grounded in Jobs-to-be-Done theory
strategic-planning-engine + thinking-systems + lyn-environmental-foresight + thinking-second-orderSystems-Aware Strategic Planning — Plans that account for feedback loops, second-order effects, and environmental forces, not just linear task lists
systematic-debugging + simota-scout + simota-specter + tob-codeql-analysis + simota-rewindDeep Bug Forensics — Not just finding bugs, but understanding their root cause, their history, their systemic patterns, and preventing their recurrence

How to use: For complex tasks, don't just pick one skill — compose multiple skills into an emergent capability that's greater than the sum of its parts.

Type 6: Tension/Dialectic

Two skills that pull in opposite directions, creating productive tension that leads to better outcomes.

Skill A (Thesis)Skill B (Antithesis)Synthesis
thinking-first-principles (simplify to essentials)thinking-systems (everything is connected)Know when to decompose and when to see the whole
lyn-focus-timeboxing (ship fast, 80/20)code-review-excellence (thoroughness, quality)Ship fast on the right things, be thorough on the critical things
creative-connections-engine (diverge, explore)meta-thinking-engine (converge, evaluate)Alternate between creative expansion and critical contraction
thinking-via-negativa (remove, simplify)pm-feature-investment (add features for ROI)Remove what doesn't matter so you can invest in what does
simota-bolt (optimize for speed)wondelai-clean-code (optimize for readability)Fast code that's also readable — the real craft
thinking-effectuation (work with what you have)thinking-first-principles (reimagine from scratch)Know when to be pragmatic and when to be revolutionary

How to use: When two skills seem to conflict, don't pick one — hold both in tension. The synthesis is usually better than either extreme.

Type 7: Recursive/Fractal

Skills that apply at multiple scales — the same skill works at the code level, the architecture level, the organization level, and the strategy level.

SkillCode ScaleArchitecture ScaleOrganization ScaleStrategy Scale
thinking-systemsFunction dependenciesService dependenciesTeam dependenciesMarket dynamics
thinking-theory-of-constraintsSlow functionBottleneck serviceOverloaded teamMarket constraint
thinking-feedback-loopsRecursive callsEvent loopsRetrospectivesCustomer feedback
thinking-first-principlesWhy this algorithm?Why this architecture?Why this org structure?Why this market?
thinking-inversionWhat breaks this function?What breaks this system?What breaks this team?What breaks this strategy?

How to use: When a skill works well at one scale, try applying it at a different scale. The same pattern often reveals insights at every level.


The Connection Discovery Protocol

When facing any problem, use this protocol to find unexpected skill connections:

Step 1: Name the problem domain. What field is this problem obviously in?

Step 2: Name three distant domains. What fields have nothing to do with this problem?

Step 3: For each distant domain, find one skill. What skill from that domain might have a structural analogy to this problem?

Step 4: Apply the distant skill's framework. Translate the distant skill's key concepts into the language of your problem.

Step 5: Evaluate the insight. Did the cross-domain application reveal something you wouldn't have seen otherwise?

This protocol is based on Koestler's bisociation principle (from creative-connections-engine) — creative breakthroughs happen when two previously unconnected frames of reference collide.


Living Map Principle

This connection map is not static. Every time you discover a new connection between skills, it strengthens the network. The more connections you find, the more powerful the entire skill library becomes. Each task is an opportunity to discover new connections that make future tasks easier.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.