
Company brain Q&A
Answer questions across your database, support tickets and docs using one linked memory
- Repository
- topoteretes/cognee
- Stars
- 33,002
- Licence
- Apache-2.0
- Files
- 1 file
What it does
It tells the agent how to load a company's SQL database, support ticket exports and document folder into a cognee memory graph, where the same person, team or project becomes a single linked node. After that, the agent can answer questions that need several sources together, such as who owns or is handling something across tickets, docs and data. It runs locally and follows the steps in the bundled cookbook.
Great for
- Engineering or ops teams who want one searchable view across their database, tickets and docs
- Developers evaluating cognee's company-memory cookbook before adapting it
- Someone who wants to try the workflow first on the built-in sample company
Skip it if
- Sensitive company data you are not ready to send to an LLM provider, since every run uses LLM credits and extraction sends content to the model (privacy risk)
- Quick one-off lookups in a single file, where a simpler search is enough
- Anyone without Python tooling and a configured cognee setup, since setup is non-trivial
Try asking
- “Connect our Postgres customers table, the Zendesk export and the docs folder, then ingest them”
- “Who owns the billing outage across tickets and the runbooks?”
- “Which team decided on the new pricing, and what tickets came after it?”
- “Run the company brain sample and ask a question that needs all three sources”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the Apache-2.0 licence and an attribution note.
company-brain-company-qa.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip company-brain-company-qa.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip company-brain-company-qa.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Start with no sources to see the sample run before connecting real data
- Use --check first to confirm setup, and fix any MISSING line the output names
- Ask follow-up questions with scripts/ask.py instead of re-ingesting, to save LLM credits
What's inside
- SKILL.md
Redistributed unmodified from topoteretes/cognee under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.