Propose a refinement (provenance pointers only). It lands in the owner's confirm queue — nothing changes until they accept. Authorized by OAuth (own model) or a contributionToken.
| Name | Required | Description | Default |
|---|---|---|---|
| items | Yes | ||
| modelId | No | ||
| contributionToken | No | Scoped propose-only key, if not connected via OAuth. |
| Name | Required | Description | Default |
|---|---|---|---|
| note | Yes | ||
| status | Yes | ||
| reviewUrl | Yes | ||
| proposalIds | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description adds important behavioral traits not present in the annotations: the queued confirmation workflow ('nothing changes until they accept') and authentication requirements (OAuth or contributionToken). It also restricts inputs to provenance pointers, which is beyond what annotations provide.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is three concise sentences, each adding distinct information: purpose, queued behavior, and authorization. It is front-loaded with the action verb and contains no redundant wording.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
The description covers purpose, queue behavior, and authorization, which is sufficient for an agent to choose the tool correctly. Given that an output schema exists and annotations are present, return values and safety profile are already available. However, a brief note on how to identify modelId would make it more complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is only 33%, and the description does not compensate. It mentions 'provenance pointers only' which adds some meaning to the items array, but modelId is completely unexplained and the structure of items is only partially clarified by the schema's nested descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a specific verb and resource: 'Propose a refinement', and adds the scope constraint 'provenance pointers only'. It clearly distinguishes itself from the sibling apply_refinement by explaining that the proposal lands in a confirm queue and nothing changes until accepted.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description gives clear context that this tool is for proposing refinements without immediate effect, contrasting with apply_refinement. It doesn't explicitly name alternatives or give when-not-to-use guidance, but the queue behavior makes the usage context clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Most tools are clearly distinct: add_hunch and suggest_question differ by confirmed vs. proposed; propose_refinement and apply_refinement differ by queued vs. direct. The only potential confusion is find_complementary_thinkers vs. find_matches, but their descriptions (famous thinkers vs. actual people) mitigate overlap.
All tool names follow a consistent verb_noun pattern (add, apply, find, propose, read, suggest), with no mixed conventions or vague verbs. The names are descriptive and predictable.
Seven tools is well within the ideal range for a focused domain. Each tool serves a distinct core operation—reading, adding, suggesting, refining, and matching—so no tool feels redundant or excessive.
The tool surface covers the primary workflows: reading a model, adding confirmed hunches, suggesting questions, proposing/applying refinements, and finding complementary thinkers or people. Minor gaps exist—no explicit update/delete hunch tool, and pending proposals are not listable—but these may be intentionally delegated to the human hub.