Start with the files and written material that are allowed to answer the target questions.
Knowledge Agent
Knowledge base chatbot for answers buried in documents
A knowledge base chatbot gives people one controlled place to ask repeat questions against the documents and knowledge you approve.
For the usual MCP path, model-token cost to the Agent creator is $0: the customer's AI client supplies the model under that customer's subscription and limits.
Early access
What is getting in the way
Policies, product notes, training material, and expert answers spread across files. People search slowly, repeat the same questions, and often use whichever copy they find first.
What changes
Organize approved resources as Agent Knowledge, keep updates owner-controlled, and publish an Agent that can search the current knowledge available to it.
Knowledge Agent
How Prostir gets you there
Choose authoritative sources
Start with the files and written material that are allowed to answer the target questions.
Build searchable knowledge
Add the resources to the Agent's knowledge boundary and review what the Agent should answer, qualify, or refuse.
Publish to the right people
Set controlled access, test common and difficult questions, and publish the Agent only when its answer boundary is clear.
Early access
What you can verify
- Agent Knowledge is built from owner-controlled resources rather than copied from a customer's private memory.
- The runtime supports searchable Markdown knowledge and can provide source-aware answers where the configured flow carries that context.
- Knowledge updates, access, tools, and publication remain controlled by the Agent owner.
A good fit if
- Your company or customers repeatedly ask factual questions covered by approved documents.
- You can identify which sources are authoritative and who should be allowed to use them.
Not the right fit if
- You expect every website, Notion workspace, recording, or private account to be crawled automatically.
- You need a guarantee that every generated sentence will include a formal citation regardless of configuration.
Questions before you decide
01Will the Agent automatically know every company document?
No. You choose and maintain the approved resources. That explicit boundary is what keeps the knowledge useful and owner-controlled.
02Is this the same as a collaborative Company workspace?
No. A standalone knowledge Agent is controlled by its creator and used by customers or employees. A Company is for active members who also collaborate on shared Tasks, Data, Files, and Knowledge.
03Who pays for model tokens, and what do they cost the Agent creator?
For the usual MCP path, model-token cost to the Agent creator is $0. When a customer uses the Agent through a supported MCP connector in ChatGPT or Claude, the model runs in that customer's AI environment under the customer's subscription and limits. The creator pays for the Prostir plan, not those model tokens. AI Credits are needed only when the creator separately chooses a Prostir-hosted model run, such as a supported channel or widget. BYOK is optional and bills the creator's own provider account. Paid external tools remain separate.
Explore related options
Early access
Start with the questions people repeat
Bring a small set of authoritative documents and the questions they should answer. We will help define a useful first knowledge Agent.