All skills
simota avatar

/compass

@35ffd55
by shingo imotasimota/agent-skills85 stars
15

Navigating the skill ecosystem and guiding onboarding. Lists agents, recommends best fit for tasks. Don't use for task execution (Nexus), agent design (Architect).

Use this Skill: https://skilld.dev/gh/simota/agent-skills/compass

This session only. Nothing lands on disk.

referencecatalog.md

≈3.7k tokens on demand. Your agent reads this file only when SKILL.md points to it.

Skill Agent Catalog

Purpose: Complete catalog of 90 global skills and 3 repository-local extensions with descriptions and triggers. Read when: You need to look up agents by category, find agents for a specific task, or provide a full listing.


How to Use This Catalog

  • To find agents by category, browse the category sections below.
  • To find agents by task, see reference/patterns.md.
  • For the most accurate, up-to-date agent count and categories, verify against the repository directory structure.

Orchestration (4 global)

Decomposes, coordinates, and parallelizes tasks.

Agent Role Trigger Code
Nexus Team orchestrator + product delivery Multi-agent coordination and scope-adaptive product/MVP delivery (deliver) No
Sherpa Task decomposition guide Break work into atomic steps under 15 minutes No
Rally Parallel orchestrator Multi-session parallel execution No
Atelier Design → implementation pipeline Integrates Vision → Muse/Frame → Forge → Artisan → Vitrine → Canvas No

Investigation (8)

Investigation, analysis, and root-cause identification. Does not write code.

Agent Role Trigger Code
Scout Bug investigation / RCA Diagnose bug causes, build reproduction steps No
Lens Codebase comprehension Structure mapping, feature discovery, data-flow tracing No
Trail Git history investigation & legacy archaeology Regression analysis, commit archaeology, implicit business-rule extraction (static-rules) No
Ripple Impact analysis Pre-change risk evaluation, blast-radius estimation No
Sweep Dead-code detection Unused files, dead code, orphaned files No
Spark New feature proposals Feature ideas leveraging existing data/logic No
Void YAGNI verification Scope cuts, complexity reduction proposals No
PDM Delivery-status navigator Planned-vs-implemented reconciliation, feature inventory, roadmap rollup No

Implementation (5)

Code implementation.

Agent Role Trigger Code
Builder Business-logic, CLI/TUI, and grammar implementation API integrations, data models, image-generation APIs, CLI/TUI tools (cli), regex / parser / DSL (grammar) Yes
Artisan Frontend implementation Production React/Vue/Svelte Yes
Forge Prototyping Fast prototypes for both frontend and backend Yes
Native Pure-native mobile implementation iOS Swift 6.3 + SwiftUI / Android Kotlin 2.4+ + Jetpack Compose (RN/Flutter/KMP/CMP out of scope) Yes
Pixel Mockup → code Pixel-accurate HTML/CSS from images Yes

Testing (5)

Test authoring and verification.

Agent Role Trigger Code
Radar Unit tests and test data Edge cases, flaky-test repair, coverage, factories / fixtures / seeds (fixtures) Yes
Voyager Cross-platform + iOS E2E Playwright/Cypress/Appium/Detox/Maestro and XCUITest/snapshots (ios) Yes
Siege Load and resilience testing Load tests, contract tests, chaos engineering Yes
Matrix Manual QA test-case authoring Systematic QA procedures (BVA, equivalence class, decision table) for TestRail/Zephyr/Xray/Qase — qa-scenario recipe (absorbed from drill) No
Canvas Test intelligence visualization Coverage heatmaps, test-shape views, mutation overlays from junit/lcov/allure/playwright artifacts (absorbed from vista; live dashboards → Pulse) Mixed

Security (5)

Security analysis and testing.

Agent Role Trigger Code
Sentinel Static security analysis Hardcoded-secret detection, SQLi prevention, dependency CVEs Mixed
Breach Red team Attack-scenario design, MITRE ATT&CK No
Probe Dynamic security testing OWASP ZAP / Burp Suite, penetration testing Mixed
Crypt Cryptographic architecture Algorithm selection, key management, E2EE, TLS configuration Mixed
Chain (optional: incident-response) Skill/plugin/MCP supply-chain audit and malware response sha256 manifests, Unicode injection, IoC campaign scans, quarantine, eradication, and credential-rotation runbooks Mixed

Review (6)

Code review and quality checks.

Agent Role Trigger Code
Judge Automated code review PR review automation, bug detection Mixed
Zen Refactoring Variable renaming, function extraction, dead-code removal Mixed
Canon (optional: legal-jp) Standards, regulatory, and legal-document compliance OWASP/WCAG/OpenAPI, SOC2/PCI-DSS/HIPAA/ISO 27001, ToS/privacy/Tokushoho review Mixed
Gauge SKILL.md audit 19-item checklist conformance No
Attest Spec-compliance verification Acceptance-criteria extraction, BDD scenario generation No
Cloak Privacy engineering PII detection, GDPR / CCPA compliance Mixed

Performance (2)

Performance optimization.

Agent Role Trigger Code
Bolt Frontend / backend optimization & continuous auto-tuning Re-render reduction, N+1 fixes, caching; profile→parameter→optimize→verify loops (absorbed from dial) Yes
Tuner DB optimization EXPLAIN ANALYZE, index recommendations, slow queries Yes

Documentation (5)

Documentation authoring, visualization, and article writing.

Agent Role Trigger Code
Scribe Technical and cross-team specifications PRD/SRS/HLD/LLD, unified L0-L4 packages, document format conversion (convert) No
Quill Code documentation JSDoc additions, README updates, fixing any types Mixed
Prose UX writing Microcopy, error messages, voice and tone No
Tome Learning material and technical publications Diff → tutorial conversion, design-decision records, articles for note / Zenn / Qiita / dev.to No
Canvas Diagramming and visualization Mermaid / ASCII / draw.io for flow, sequence, ER diagrams Mixed

Architecture (4)

System design and structure.

Agent Role Trigger Code
Atlas Architecture analysis Dependency analysis, ADR / RFC authoring Mixed
Schema DB and tenant architecture Normalization, migrations, ER diagrams, tenant isolation/RLS/routing (tenant) Mixed
Gateway API design and messaging integration OpenAPI generation, versioning, chat adapters / webhooks / WebSocket (messaging) Mixed
Grove Human/LLM repository structure Directory/docs layout plus LLM navigation and prompt-cache topology (llm) Mixed

UX/Design (11)

UI/UX design and improvement.

Agent Role Trigger Code
Vision Creative direction UI/UX direction decisions No
Palette Usability improvement Cognitive-load reduction, a11y compliance Mixed
Echo Persona walkthrough + synthetic demand Usability, feature requests, JTBD, 5 Whys, opportunity trees (demand) No
Flow Animation implementation CSS / JS animation, transitions Yes
Muse Design tokens Token architecture, dark mode Mixed
Vitrine Storybook Story authoring, Visual Regression Mixed
Field User research Interview design, persona creation No
Trace Session-replay analysis Behavioral pattern extraction, UX issue discovery No
Cast Persona casting Persona generation, management, sync No
Funnel Landing-page construction LP conversion and premium nine-stage production (premium) Mixed
Voice User-feedback analysis NPS design, review analysis, sentiment analysis No

DevOps (5)

Infrastructure, CI/CD, and operations.

Agent Role Trigger Code
Gear Dependency management / CI/CD / GHA Build errors, pipelines, advanced GitHub Actions (gha) Yes
Scaffold Infrastructure provisioning Terraform / Docker Compose design Yes
Beacon Observability and reliability SLO / SLI design, alert strategy Mixed
Launch Release management and PR reporting Versioning, CHANGELOG, rollback, weekly/monthly reports and engineering metrics Mixed
Ledger FinOps Cloud cost optimization, RI / SP recommendations No

Modernization (2)

Migration and modernization.

Agent Role Trigger Code
Shift Migration, upgrade & modernization orchestrator Framework / library / DB migration; deprecated-library detection (detect); native-API replacement (modernize); technology radar (radar) — absorbed from horizon Mixed
Port Web → Native porting design Blueprint from Web SPA / SSR / PWA to iOS Swift/SwiftUI + Android Kotlin/Compose pure-native (parity matrix, phased roadmap) No

Growth (1)

Growth tactics and branding.

Agent Role Trigger Code
Growth SEO / CRO / GEO and retention Meta / OGP / JSON-LD, CTA optimization, re-engagement and loyalty design (retention) Mixed

Analytics (3)

Metrics, experimentation, and combinatorial analysis.

Agent Role Trigger Code
Pulse KPI design North Star Metric, funnel analysis Mixed
Experiment A/B test design Hypothesis documentation, sample-size calculation Mixed
Matrix Combinatorial analysis Combination-explosion control, minimum coverage No

Git/PR (1)

Version-control workflow.

Agent Role Trigger Code
Guardian PR management Change classification, granularity recommendations, strategy No

Browser (1)

Browser automation and asset acquisition.

Agent Role Trigger Code
Vector Browser automation and crawl architecture Task completion via Playwright / DevTools, crawler topology and frontier design (crawl) Yes

Data (1)

Data pipelines and conversion.

Agent Role Trigger Code
Stream ETL / ELT pipelines Kafka / Airflow / dbt design Mixed

Strategy (2)

Business strategy and decision-making. Does not write code.

Agent Role Trigger Code
Compete Competitive research and personal branding Feature matrices, product positioning, GitHub / LinkedIn / blog / conference positioning No
Rank Prioritization ICE / RICE / WSJF / MoSCoW scoring No

Incident (2)

Incident response.

Agent Role Trigger Code
Triage Incident first response Impact identification, recovery procedure, postmortem No
Mend Automated remediation Runbook execution, staged verification, rollback Mixed

Communication (1 agent + Scribe recipe)

Coordination and communication.

Agent Role Trigger Code
Scribe (unified) Spec alignment Cross-team Business / Dev / Design specifications No

Meta / Tooling (5 global)

Manages and evolves the ecosystem itself.

Agent Role Trigger Code
Architect (skill-meta) Skill design New agent design, gap analysis No
Sigil (skill-meta) Project operating-layer design Tailored skills plus recipe/workflow/routing-map blueprints (blueprint) No
Hone (optional: ai-cli-admin) AI CLI configuration, hooks, and personal environment CLI audits, PreToolUse / PostToolUse / Stop hooks, dotfiles and shell/editor config (env), macOS AppleScript / JXA automation (automate) Mixed
Compass Skill navigator Skill guidance, onboarding No
Prune (skill-meta) Skill ecosystem cleanup Overlap audit, merge-candidate detection, sunset proposals No

Creative / Media (1)

Media and creative generation.

Agent Role Trigger Code
Ink SVG icons Icon systems, sprite construction Yes

AI / ML (3)

AI design and thinking support.

Agent Role Trigger Code
Oracle AI / ML design Prompt engineering, RAG design, evaluation No
Chisel Prompt → executable spec Ambiguity detection, criterion translation, role decomposition; Nexus SPECIFY briefs No
Flux Thinking refraction and ideation Assumption challenges, perspective shifts, multi-turn Expand / Propose / Evaluate / Subtract dialogue (ideate) No

Other Specialists

Specialized skills that don't fit the categories above.

Agent Role Trigger Code
Polyglot i18n / l10n Multilingual support, RTL handling Mixed
Weave Workflow and schedule design State machines, Saga patterns, cron / timezone / business calendar / backfill (schedule) Mixed
Omen Pre-mortem analysis Failure-scenario enumeration, RPN scoring No
Seek Search-engine design Full-text search, vector search, RAG Mixed
Vigil Detection engineering Sigma / YARA rule design Mixed
Magi Multi-perspective deliberation, advisory, and strategy simulation Architecture arbitration, Go / No-Go, founder coaching, named-figure lenses, business scenario simulation (simulate) No
Saga Narrative design Customer-experience storytelling No
Cue Video script and demo production Product videos, storyboards, Playwright-generated demos Mixed
Stage Slide generation Marp / reveal.js / Slidev Mixed
Frame Figma → code bridge Design context extraction No

Project-local extensions (3)

Available only inside this repository. Check _common/PROJECT_LOCAL_SKILLS.md before routing; use its fallback when unavailable.

Agent Role Trigger Code
Orbit Autonomous-loop design Script generation for this repository's nexus-autoloop Mixed
Lore Knowledge curation .agents/*.md synthesis into METAPATTERNS.md No
Darwin Ecosystem evolution Fitness evaluation persisted to .agents/ECOSYSTEM.md No

Source: SKILL.md on GitHub

1 warning13d4 checks · Risk SAFE
  • Gen Agent Trust Hub13d

    The Compass skill functions as a navigator and onboarding guide for a large ecosystem of AI agents. It recommends the most appropriate agents for a given task by analyzing the project's technology stack through repository configuration files (like package.json). Although it reads external data which creates a surface for indirect prompt injection, the skill's capabilities are strictly limited to providing guidance and navigation, with no access to network exfiltration or arbitrary code execution, resulting in a low overall risk.

  • Socket13d

    No alerts

  • Snyk13d

    Risk: MEDIUM · 1 issue

  • ZeroLeaks5mo

    Score: 93/100 · 2 sections analyzed

Signed by skilld at 35ffd55. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 2 days ago.

Activeupdated 2 weeks ago

README badge

README badge for simota/agent-skills/compass