fusengine/agents/research

<objective> Research is the technical-research methodology behind the research-expert agent, covering three workflows depending on the question: a standard query (Sequential Thinking decomposition -> Context7 resolve + docs -> Exa supplement -> synthesized answer with sources), a complex investigation (multi-hypothesis Sequential Thinking -> Exa deep researcher -> cross-check against Context7 -> comprehensive report), and a technology-trends scan (Exa web scan + code patterns + company research -> Sequential Thinking analysis -> recommendations).

Source priority is always official docs (Context7) first, recent tutorials (Exa, <6 months) second, older content only with version verification. It runs in a forked context via the research-expert agent. </objective>

Session: ${CLAUDE_SESSION_ID}

Research Topic: $ARGUMENTS

Research Skill

Research Workflows

Standard Query

1. THINK → Sequential Thinking decomposition
2. RESOLVE → Context7 resolve-library-id
3. DOCUMENT → Context7 query-docs (5000-10000 tokens)
4. SUPPLEMENT → Exa code context search
5. SYNTHESIZE → Structured answer with sources

Complex Investigation

1. DEEP THINK → Multi-hypothesis Sequential Thinking
2. DEEP RESEARCH → Exa deep researcher (45s-2min)
3. MONITOR → Check status until completed
4. VALIDATE → Cross-check Context7 official sources
5. REPORT → Comprehensive solution

Technology Trends

1. WEB SCAN → Exa search latest developments
2. CODE PATTERNS → Exa code context for practices
3. ECOSYSTEM → Company research for key players
4. ANALYSIS → Sequential Thinking for implications
5. RECOMMENDATIONS → Actionable insights

Multi-Source Synthesis

Parallelization:

• Run resolve-library-id + web_search_exa simultaneously
• Launch multiple Exa searches concurrently
• Execute Context7 docs + Exa code search in parallel

Source Priority:

1. Official documentation (Context7)
2. Recent tutorials (Exa, <6 months)
3. Older content (with version verification)

Forbidden Behaviors

• ❌ Guess library IDs without resolve-library-id
• ❌ Start deep researcher without checking completion
• ❌ Mix opinions with facts without distinction
• ❌ Provide code without version verification
• ❌ Ignore WebFetch redirects
• ❌ Recommend without citing sources
• ❌ Skip Sequential Thinking for multi-step problems

Detailed References (Load on Demand)

• references/tool-usage.md — Load when calling Context7, Exa, or Sequential Thinking (exact call patterns, models, and per-tool error handling)
• references/response-format.md — Load when writing the final research report