agentleFS
equinorequinor/fusion-skills/skills/.system/fusion-research/SKILL.md

fusion-research

Skill2 starsChanged 2 months agoMITView on GitHub

Source-backed research orchestrator for the Fusion ecosystem. Routes to the correct research agent based on question type. Returns source-backed evidence only; will not invent Framework behavior, component APIs, skill catalog relationships, platform guidance, or backend implementation details. USE FOR: any research question needing source-backed evidence about Fusion Framework APIs, EDS components, the Fusion skill catalog, Fusion platform documentation, or backend service implementation. DO NOT USE FOR: implementing code changes, installing or editing skills, MCP setup or troubleshooting, or inventing Fusion behavior without evidence.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 7 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when a user needs a source-backed answer about the Fusion ecosystem.

    line 37
  2. Distinctive:

    Do not use this skill for:

    line 57
  3. Distinctive:

    Before classifying, you may use the domain-specific follow-up questions to narrow the user's exact need:

    line 68
  4. Distinctive:

    If the question spans multiple domains, answer each domain separately using the appropriate agent in sequence.

    line 88
  5. Distinctive:

    If the runtime supports skill-local agents, invoke the agent directly.

    line 98
  6. Distinctive:

    Use the structure in assets/source-backed-answer-template.md:

    line 102
  7. Distinctive:

    keep refining indefinitely — state uncertainty after one refinement pass per agent

    line 130

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Fusion Research

When to use

Use this skill when a user needs a source-backed answer about the Fusion ecosystem.

Typical triggers:

  • "Research this Fusion Framework hook and show me the supporting source."
  • "What props does this EDS component accept?"
  • "Which skill handles X in a Fusion workflow?"
  • "Find a source-backed example for…"
  • "Which Fusion package or module owns this API?"
  • "Is there an EDS design token for this value?"
  • "How can I persist user preferences in a Fusion app?"
  • "What's the recommended pattern for X in Fusion?"
  • "Show me a Fusion Framework example for building X."
  • "What are the design tokens for color and spacing in EDS?"
  • "How does Fusion handle X at the platform level?"
  • "What is the Fusion platform guidance on Y?"
  • "How do I onboard to Fusion?"
  • "What are the guidelines for Z in the Fusion platform?"

When not to use

Do not use this skill for:

  • implementing code changes — use fusion-app-react-dev or the relevant dev skill; if the research is for a specific app you are actively building, use fusion-app-react-dev first — it will call this skill when needed
  • finding, installing, updating, or removing skills — use fusion-discover-skills; use this skill only to understand a skill's scope, relationships, or catalog fit — not to discover what to install
  • creating or editing skill files — use fusion-skill-authoring for authoring-time research
  • Fusion MCP installation or troubleshooting — use fusion-mcp; once MCP is running, return here for research
  • pure conceptual Fusion questions with no specific artifact name (hook, component, package, token, or platform topic) — ask the user for a concrete artifact reference or platform topic first

Instructions

Step 0 — Sharpen scope (optional)

Before classifying, you may use the domain-specific follow-up questions to narrow the user's exact need:

  • Framework scope: assets/framework.follow-up.md
  • EDS scope: assets/eds.follow-up.md
  • Skills catalog scope: assets/skills.follow-up.md
  • Docs scope: assets/docs.follow-up.md

Skip this step if the question already names a specific artifact (hook, component, package, token, or skill).

Step 1 — Classify the research question

Determine which domain the question belongs to:

DomainIndicatorsAgent
FrameworkFusion Framework hooks, packages, modules, TypeScript APIs, cookbook examplesagents/framework.agent.md
EDSEDS component props, usage examples, accessibility, design tokensagents/eds.agent.md
SkillsSkill catalog lookup, scope boundaries, companion/orchestrator relationshipsagents/skills.agent.md
DocsFusion platform concepts, onboarding, platform operations, governance, non-implementation guidanceagents/docs.agent.md
Backend CodeC# service implementations, interfaces, CQRS patterns, authorization, validation, cross-service APIsagents/backend-code.agent.md

If the question spans multiple domains, answer each domain separately using the appropriate agent in sequence.

Step 2 — Dispatch to the correct agent

  • Framework questions → follow agents/framework.agent.md.
  • EDS questions → follow agents/eds.agent.md.
  • Skills questions → follow agents/skills.agent.md.
  • Docs questions → follow agents/docs.agent.md.
  • Backend code questions → follow agents/backend-code.agent.md.

If the runtime supports skill-local agents, invoke the agent directly. Otherwise, apply the agent's instructions inline.

Step 3 — Return the source-backed answer

Use the structure in assets/source-backed-answer-template.md:

  • State which domain and agent was used.
  • Include one to three source-backed evidence bullets.
  • End with any remaining assumptions, uncertainty, or the next verification step.

Research agents

This skill includes five research agents in agents/. Each covers one Fusion research domain with its own query patterns and evidence checklist.

  • agents/framework.agent.md — source-backed answers about Fusion Framework hooks, packages, modules, and cookbook examples. Uses mcp_fusion_search_framework.
  • agents/eds.agent.md — source-backed answers about EDS component props, usage, accessibility, and design tokens. Uses mcp_fusion_search_eds.
  • agents/skills.agent.md — source-backed answers about the Fusion skill catalog: what skills exist, their scope, and how they relate. Uses mcp_fusion_search_skills.
  • agents/docs.agent.md — source-backed answers about Fusion platform concepts, onboarding, operations, and governance. Uses mcp_fusion_search_docs.
  • agents/backend-code.agent.md — source-backed answers about Fusion backend service implementations: C# services, interfaces, CQRS patterns, authorization, and cross-service integrations. Uses mcp_fusion_search_backend_code.

Assets

  • assets/source-backed-answer-template.md
  • assets/framework.follow-up.md — pre-dispatch scope questions for Framework research
  • assets/eds.follow-up.md — pre-dispatch scope questions for EDS research
  • assets/skills.follow-up.md — pre-dispatch scope questions for skills catalog research
  • assets/docs.follow-up.md — pre-dispatch scope questions for Docs platform research

Safety & constraints

Never:

  • invent Fusion Framework hooks, EDS props, or skill catalog entries
  • claim evidence exists when MCP is unavailable or results are weak
  • keep refining indefinitely — state uncertainty after one refinement pass per agent
  • implement code changes or mutate repositories during a research flow

Always:

  • classify the question before choosing an agent
  • capture the source path and supporting excerpt before finalizing any claim
  • state explicitly which agent and source backed each part of the answer

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.