company-brain

Organizes team knowledge into a structured markdown vault with capture, compile, query, and review modes.

Updated Aug 13, 2026
One-click install
npx skills add https://github.com/imMamdouhaboammar/marketing-skills --skill company-brain-immamdouhaboammar
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: company-brain
Source: https://github.com/imMamdouhaboammar/marketing-skills/tree/main/skills/company-brain
Command: npx skills add https://github.com/imMamdouhaboammar/marketing-skills --skill company-brain-immamdouhaboammar

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Teams lose critical context scattered across call transcripts, Slack threads, emails, and CRM records, making it impossible to answer questions like "what do we know about this customer" without digging through disconnected tools. ## Core Features & Use Cases - Structured Capture & Compilation: Routes incoming content into dedicated directories (people/, companies/, meetings/, sops/, decisions/, customer-language/, recurring-questions/, sales-objections/) and compiles them into an interlinked wiki with multi-author attribution. - Trust-Weighted Querying: Answers questions from the corpus while respecting trust levels (unreviewed/verified/deprecated/superseded) and sensitivity tags (internal/leadership/confidential), flagging low-confidence answers. - Review & Lint Workflows: Provides a human triage queue to verify, deprecate, or supersede stale captures, plus lint checks for orphans, stale SOPs, unanswered objections, and review backlogs. - Use Case: A sales team captures call transcripts from Fathom into meetings/, compiles them into wiki pages about active deals, then queries "what objections has Acme raised" before a renewal call and gets an answer citing verified sources with author attribution. ## Quick Start Ask the agent to capture this meeting note into the team brain, then query what the team knows about a specific customer.

Frequently Asked Questions about company-brain

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a shared team knowledge base for AI agents?

Create a markdown vault with structured directories for people, companies, meetings, SOPs, and decisions, then capture content with author and trust metadata. Compile captures into interlinked wiki pages and query the corpus with trust-weighted answers.

How do I capture meeting transcripts into a knowledge base?

Save transcripts to a meetings/ directory using the naming pattern YYYY-MM-DD-company-slug.md with source, author, and captured date metadata. Auto-sync from Fathom, Gong, or Granola via webhooks, or capture manually until the team proves consistent usage.

What is the difference between company-brain and second-brain?

Company-brain is team-scope with structured directories for people, companies, and meetings plus multi-author attribution and trust levels. Second-brain is personal-scope with a flat type-prefixed raw directory designed for individual use.

Can I sync CRM data like HubSpot or Attio into a markdown vault?

Yes, CRM sync pulls deal state, contact info, and opportunity size via API on a nightly schedule, updating companies/ and people/ files in place. Sync jobs must be idempotent, deduplicating by source ID or timestamp before appending.

How do I prevent stale or wrong information from polluting AI answers?

Use a trust field with unreviewed, verified, deprecated, and superseded values instead of deleting files. Run a regular review pass to triage unreviewed captures, and deprecated content is excluded from queries while history is preserved.

When should a team not auto-sync data into a knowledge base?

Avoid auto-sync for confidential content, compliance-restricted industries, rate-limited APIs, or teams that have not proven manual capture habits. Start with 4-6 weeks of manual capture before wiring automation.