Takibi BaseVerified tool

Takibi Base is a shared knowledge base for AI agents. Teams can organize documents, set agent-specific document permissions, and retrieve cited source passages when agents ask questions.

What it does

Stores documents in projects and folders.

• Converts uploaded files to Markdown so agents can search them.
• Lets agents query permitted sources through the CLI or API and receive matching passages with citations.
• Returns no passages when it finds no matching evidence.

Key features

• Profiles that hold keys and document-access permissions for individual tools or agents.
• Folder-level access controls, with project-level access available for broader permissions.
• Key management, including individual key rotation or revocation and the ability to pause a Profile.
• Document status tracking for indexed, converting, and quarantined files.
• Original-file downloads, text extraction, and project export in Open Knowledge Format (OKF).
• Query responses that include passages, citations, and an evidence support score.
• Profile activity views for reviewing requests from keys and changes to a Profile.
• Private, Profile-specific notes with a fixed 30-day lifetime that can be reviewed before being added to source documents. What makes it different Takibi Base emphasizes source-grounded retrieval: agents receive the exact source text and a citation for each returned passage. Its Profile model combines keys with explicit document permissions, helping teams control which projects and folders each agent can read.