company-research
npx machina-cli add skill simonstrumse/vibelabs-skills/company-research --openclawCompany Research
Tool Restriction (Critical)
ONLY use web_search_advanced from Exa. Do NOT use web_search_exa or any other Exa tools.
Token Isolation (Critical)
Never run Exa searches in main context. Always spawn Task agents:
- Agent runs Exa search internally
- Agent processes results using LLM intelligence
- Agent returns only distilled output (compact JSON or brief markdown)
- Main context stays clean regardless of search volume
Dynamic Tuning
No hardcoded numResults. Tune to user intent:
- User says "a few" → 10-20
- User says "comprehensive" → 50-100
- User specifies number → match it
- Ambiguous? Ask: "How many companies would you like?"
Query Variation
Exa returns different results for different phrasings. For coverage:
- Generate 2-3 query variations
- Run in parallel
- Merge and deduplicate
Categories
Use appropriate Exa category:
- company → homepages, gargantuan amount of metadata such as headcount, location, funding, revenue
- news → press coverage
- tweet → social presence
- people → LinkedIn profiles (public data)
Public LinkedIn via Exa: category "people", no other filters Auth-required LinkedIn → use Claude in Chrome browser fallback
Browser Fallback
Auto-fallback to Claude in Chrome when:
- Exa returns insufficient results
- Content is auth-gated
- Dynamic pages need JavaScript
Models
- haiku: fast extraction (listing, discovery)
- opus: synthesis, analysis, browser automation
Source
git clone https://github.com/simonstrumse/vibelabs-skills/blob/main/skills/company-research/SKILL.mdView on GitHub Overview
Company-research taps Exa search to surface company data, competitors, news, tweets, financials, and LinkedIn profiles, then compiles them into usable company lists. It supports market research, competitive analysis, and due diligence by delivering structured insights quickly.
How This Skill Works
It runs 2-3 parallel Exa queries (varying phrasings) to cover homepages, news, social, and people data, then an agent distills results into compact JSON or brief Markdown. The main context deduplicates, links related entities, and builds organized company lists for your analysis.
When to Use It
- Entering a new market and mapping competitors
- Benchmarking financials, funding, and headcount across peers
- Finding potential partners, vendors, or acquisition targets
- Tracking breaking news and social chatter about a company
- Due diligence for investments or partnerships
Quick Start
- Step 1: Define objective and target scope (how many companies)
- Step 2: Run 2-3 Exa query variations in parallel
- Step 3: Review the compact JSON/summary returned by the agent and build your list
Best Practices
- Define scope and number of companies before running queries
- Use 2-3 query variations and merge results
- Verify high-impact data (financials, leadership) via primary sources
- Leverage LinkedIn data for public profiles but be mindful of access limits
- Organize output into deduplicated lists with unique IDs
Example Use Cases
- Map out competitors of a target through homepages, news, and social signals
- Assemble a vendor shortlist with financial health and locations
- Monitor a company's media coverage and leadership changes
- Profile LinkedIn presence of key executives for outreach
- Create a comprehensive list of market players in a niche
