Answer questions from local document collections; responses require valid source IDs and exact quotes, so answers stay grounded in retrieved text.
Retrieve then answer locally. Returns no answer without valid source IDs and exact source quotes; semantic correctness still needs review.
| Name | Required | Description | Default |
|---|---|---|---|
| model | Yes | ||
| query | Yes | ||
| top_k | No | ||
| min_score | No | ||
| collection | Yes | ||
| embedding_model | Yes |
Changes observed during successful MCP inspections.
v2.1.0Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With annotations present (readOnlyHint=false, openWorldHint=false, destructiveHint=false), the description adds genuinely non-obvious behavior: it will return NO answer unless valid source IDs and exact source quotes exist, and semantic correctness still requires human review. That prevents an agent from treating an empty result as a failure and flags the trust boundary. It stops short of explaining why readOnlyHint is false for what sounds like a read operation.
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?
Two tight sentences, front-loaded with the core behavior and the failure condition. No filler, though the extreme compression trades away detail an agent would need.
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?
For a 6-parameter, 4-required tool with no output schema and no annotation coverage of the parameters, the description is far too thin: it omits parameter meanings, required inputs, and return shape. The behavioral caveats are useful but insufficient for correct invocation.
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?
Six parameters with 0% schema description coverage, so the description carries the full burden — yet it explains none of them (collection, query, model, embedding_model, top_k, min_score). An agent cannot know what a 'collection' or 'min_score' means here from either structured fields or the text.
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?
States the core activity (retrieve then answer locally) which implies RAG question answering, and is distinguishable from lm_rag_search/lm_rag_index by the 'answer' verb. However it never names what is being queried (a collection) or explicitly contrasts with lm_rag_search, leaving the purpose only partially pinned down.
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?
There is no statement of when to use this tool versus lm_rag_search, lm_rag_index, or lm_chat, nor any prerequisites. The reader must infer usage from the name and the fragmentary description.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.