Say what is wrong with what a request produced, and have it reworked.
Status must be "done". `feedback_md` is 1..2000 chars. Moves the thread to
refine_pending; poll get_request_status to see the revised items.
| Name | Required | Description | Default |
|---|---|---|---|
| thread_id | Yes | ||
| project_id | Yes | ||
| feedback_md | Yes |
| Name | Required | Description | Default |
|---|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations are all false, so the description carries the burden. It discloses the state transition (moves thread to refine_pending), the input constraint (feedback_md 1..2000 chars), and the precondition (status done). This is meaningful behavioral context beyond the annotations, though it doesn't cover reversibility or error handling.
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 two sentences: the first states the purpose, the second packs the key constraints and state transition. No filler words. Every sentence earns its place, and the structure front-loads the primary intent.
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?
Given the presence of an output schema (which handles return values), the description covers the essential behavioral contract: when to use (status done), what to provide (feedback_md), what happens (thread moves to refine_pending), and how to observe results (poll get_request_status). This is complete for a mutation tool of moderate complexity.
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 0%, so the description must compensate. It does explain feedback_md's length constraint and purpose ('what is wrong'), but project_id and thread_id are left to their names, which are self-explanatory but not explicitly described. This is partial compensation, missing the opportunity to clarify the role of each parameter beyond the schema's bare definitions.
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 clearly states the tool's function: 'Say what is wrong with what a request produced, and have it reworked.' This is a specific verb+resource combination that distinguishes it from sibling tools like answer_request (answer a request) and submit_request (submit a new request). The context of reworking an existing request's output is unambiguous.
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 provides explicit usage context: the status must be 'done' before calling, and it instructs to poll get_request_status afterward to see revised items. It doesn't explicitly name alternatives or exclusions, but the precondition and next-step guidance effectively differentiate when to use this tool among the request-related siblings.
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.
Every tool targets a distinct resource and action duo, even within clusters like request handling or security reviews. The get_ vs run_ pairs are clearly separated, and descriptions explicitly contrast confusing alternatives such as archive_project vs delete_project.
The set overwhelmingly follows verb_noun snake_case (submit_request, list_projects, resolve_escalation). The one visible deviation is project_status, which breaks the get_/pattern, and signup is a single-word verb instead of sign_up.
37 tools is well above the 25+ threshold and spans auth, billing, project lifecycle, roadmap, escalations, product documents, and multiple review types. Most tools earn their place, but the surface is too large for one server and would be more coherent split into focused servers.
The domain coverage is broad: full project lifecycle, request intake/refinement, roadmap manipulation, escalation handling, product doc read/write, and security/legal review flows. Minor gaps exist, most notably no dedicated task-listing or task-update tool, but agents can work around these via project_status and list_escalations.