gograph_coverage

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint=true and idempotentHint=true. The description goes well beyond that: it discloses that the server refreshes the graph first, explains how exact versus possible results are determined via static vs CHA/parser-only dispatch edges, and specifies ambiguity handling. It explicitly states 'Read-only; no side effects' and distinguishes static attribution from runtime or branch coverage, which is critical behavioral context for an AI agent.

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 well structured and information-dense, using clear sections for when to use, not to use, and return value. Every sentence adds behavioral or semantic detail. It is longer than minimal, but the tool involves non-trivial precision semantics and ambiguity handling, so the length is justified. The most important statement about transitive reachability is front-loaded.

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 is complete for a read-only analysis tool with no output schema. It covers the core behavior, the degradation semantics, ambiguity resolution, retry strategies, exact-mode filtering, the optional package qualifier, and a summary of what the return payload contains. There are no obvious missing behavioral details an agent would need to invoke the tool correctly.

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 input schema already has 100% description coverage for all three parameters. The tool description still adds meaningful semantic context: 'test' is clarified as exact name or canonical stable ID with guidance to use matched_tests/gograph_identity for retries; 'package' is defined as resolving the uncommon in-package versus external foo_test collision; 'exact_only' is connected to omitting possible results. This expands beyond the schema 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 opens with a specific verb-resource statement: 'Return the transitive set of production functions and methods statically reachable from one test function.' It clearly defines the input (a test function) and the output relationship, and it disambiguates from siblings like gograph_tests, gograph_untested, and gograph_callees by emphasizing static reachability from a single test rather than coverage claims or general call graphs.

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 explicitly provides WHEN TO USE ('map one test to the product symbols it structurally exercises') and NOT TO USE ('To claim execution or branch coverage') sections. It also gives concrete conditional guidance for ambiguous cases: same-named tests return status=ambiguous, and the agent should retry with the stable test ID from matched_tests or gograph_identity. The package qualifier is framed precisely as resolving only the in-package versus external foo_test collision.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.