get_subject_classificationRead the current classification state and its decision audit. Confirmed classifications are locked and must not be routinely reassessed.
| Name | Required | Description | Default |
|---|---|---|---|
| subject_id | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint, and destructiveHint, so the safety profile is covered. The description adds meaningful behavioral context beyond annotations: it returns not only the classification state but also the decision audit, and it warns that confirmed classifications are locked against routine reassessment.
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 short sentences with no filler. The primary verb and resource are front-loaded, and the policy warning is presented distinctly after the functional description.
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 one-parameter read operation with rich annotations, the description covers the essential information: what is returned (state and audit) and a key policy constraint (locked confirmed classifications). It does not detail error or not-found behavior, but that is a minor gap given the simplicity of the tool.
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?
The schema contains a single subject_id parameter with 0% description coverage, and the description does not mention the parameter at all. The parameter name is somewhat self-explanatory, but the description adds no semantic guidance about what subject_id refers to or how it relates to the classification state.
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 ('Read') and specific resources ('current classification state' and 'decision audit'), making the tool's purpose immediately clear. It also differentiates from siblings like propose_subject_reclassification and reopen_subject_classification by noting that confirmed classifications are locked and not routinely reassessed.
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 clearly indicates when to use this tool: when you need the current settled classification state and its audit trail. It also provides an implicit exclusion by warning that confirmed classifications must not be routinely reassessed, though it does not explicitly name alternative tools.
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 separated by resource and action, but the classification cluster (affirm_subject_classification, propose_subject_reclassification, reopen_subject_classification, get_subject_classification) and the three resolve_subject* tools require close reading to distinguish. Long descriptions help, but an agent could easily pick the wrong member of those clusters.
The dominant pattern is verb_noun in snake_case (create_deliberation, list_reviews_by_visibility, resolve_subject_type, set_review_visibility). Minor deviations such as bare 'fetch' and 'search' and mixed verbs like affirm vs propose vs reopen are readable and do not break the convention.
34 tools substantially exceeds the 25+ threshold even for a server with multiple subdomains. The many classification, type, and location variants add cognitive load, and several could be consolidated or hidden behind a smaller workflow-oriented surface.
The set covers the main lifecycles: review save/fetch/assess/delete/visibility, subject resolution/enrichment/correction, classification with voting/reopening, deliberations, locations, and vocabulary management. Minor gaps remain, such as no direct full-subject getter and no update to review text, but these appear intentional in an immutable-review design.