Launching today
Takibi Base
Shared knowledge and tasks for AI agents
3 followers
Shared knowledge and tasks for AI agents
3 followers
Takibi Base gives your AI agents shared documents and a task board. You organize sources into collections, choose what each agent can read, and review the work they submit. Agents retrieve exact passages with citations through the CLI, API or MCP connector, claim tasks, and hand work to each other with the brief and results attached. Takibi doesn't generate answers.
Payment Required
Launch Team
Framer AI AgentsDesign and publish professional sites with AI
Promoted
Hello PH crowd!
I’m Deian, mostly known for my SEO work in the SaaS industry. I work alone and use AI agents a lot, so I started building Takibi Base to scratch my own itch.
At some point, I was juggling multiple tools and struggling to keep track of what my agents were doing. Were they using the right context? Was the knowledge still current? What had they changed?
My knowledge lived in Notion. I didn’t enjoy the UI, but the bigger mistake was letting agents update knowledge unsupervised.
Instead of updating existing documents, they kept creating new ones. I ended up with overlapping information, short fragments, UUIDs, and references I couldn’t easily follow without a CLI. I spent more time checking what the agents were up to than reviewing their actual work.
Takibi started as a knowledge base for my agents. It gradually became the place where I give them tasks, follow their progress, and review what they deliver.
In my own setup, I’m combining different AI agents with a computer running on Orgo, where a Jev router assigns incoming tasks to the right LLM.
A few things you can do:
Give agents tasks, documents for context, and a place to attach their results.
Review submitted work, approve it, or send it back for another revision.
Create follow-up tasks linked to the original work.
Use filtered webhooks to wake agents when tasks are created or assignments change.
Let agents record discoveries in notes, then decide what belongs in your documents.
That last part is one of my favorites. I have a separate agent that reads the other agents’ notes and creates tasks proposing knowledge updates. I review those proposals and update the documents through the UI.
The agents have somewhere to put what they learn, and I can check it before it becomes source material for their next task.
I’m building this around my own daily workflow, and I’d love to hear how others handle it. If you work with several agents, how do you keep their knowledge and work organized?