Answers questions only from facts a person approved, each fact with a link. Read public knowledge stores or create your own.
Scored across 7 tools
The step-oriented tools (startStep, getStep, replyToPrompt) have reasonably distinct roles, but `ask` overlaps directly with the `ask` step type in `startStep` and also returns step state, making selection ambiguous. Other tools like createStore, getManual, and listFacts are clear.
Most names follow a camelCase verb_noun or verb_object pattern (createStore, getManual, getStep, listFacts, replyToPrompt, startStep), but `ask` is a bare verb and breaks the predictable pattern.
Seven tools is well-scoped for a fact-store workflow: one creation tool, one manual, one listing tool, and four step-lifecycle tools. No tool seems extraneous or missing for the stated protocol.
The core add/list/ask/remove lifecycle is present via steps, but there is no direct update/edit fact operation, no get-specific-fact tool, no search/filter, and no store deletion. These are notable gaps an agent must work around, often by removing and re-adding facts.
Ask a question; the store answers only from its facts, each claim linked to its fact, and checks the answer against them. Stores nothing. Returns the step's state: waiting (a prompt to do and answer with replyToPrompt), running (call getStep), done (do what next says) or failed.
Inputs: question (required), store.
A new, empty FactoGraph store of the person's own: its private link, which these tools take as store. Only after the person said yes to having one. Nothing is stored until the link is first opened. The link is the store's key: give it to them once, and never put it anywhere public.
Inputs: person_agreed (required).
The store's manual - how to run its steps and cite its facts - and what it holds: how many facts, from which sources. Read it before the store's other tools, and follow it.
Inputs: store.
Where a running step is now: its next prompt, or its result.
Inputs: run_id (required), store.
Every fact in the store, numbered, each number a link to the fact. Cite a fact as its number with its link, F3.
Inputs: store.
Answer the prompt a step is waiting on - do the model work in its system and messages first. Keep answering until the step is done or failed.
Inputs: reply (required), run_id (required), call_id (required), store.
Start a step - propose, add, documents, ask or remove - in a private store, with its JSON body, as the manual says. propose stores nothing; add stores only facts the person approved, with propose's proposal_id and their reply as confirmation; remove deletes only with confirm true, after the person confirmed. Returns the step's state: waiting (a prompt to do and answer with replyToPrompt), running (call getStep), done (do what next says) or failed.
Inputs: body (required), step (required), store (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Add one secure layer between your agents and this server.