ask_gemini

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

With no annotations, the description carries the full behavioral burden and it delivers: single-turn semantics, the binary_missing failure mode when the agy CLI is unavailable, and explicit refusals for offensive-security and non-software domains. It even clarifies in-scope domains (neuroscience, cognitive science, AI/ML, CS) so the agent can distinguish refusals from valid questions.

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 front-loaded with purpose, then usage, then failure/refusal behavior, then sibling routing — no redundant filler. Each sentence contributes distinct decision-relevant information for an agent choosing and invoking the tool.

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 single-turn LLM query tool with no output schema, the description covers needed dependencies, refusals, and scope, and the schema covers parameter semantics. The only minor gap is that it never names the response shape, but that is reasonably inferable and partially disclosed through the binary_missing error mention.

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 documents all three parameters at 100% coverage, including the context_ref key behavior. The tool description adds marginal value by advising to place a snippet or file path in context when a question references existing code and by emphasizing the question must be software/engineering-specific.

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 first clause names a specific verb ('Ask'), a specific model ('Gemini 3.1 Pro'), and the invocation path ('local agy CLI'), and narrows the subject to software/engineering work. It also distinguishes itself from Fable-bound siblings and from the other ask_* tools by listing the exact kinds of questions it handles.

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 states when to use the tool (broad/conceptual software engineering questions, brainstorming), when to include context (questions about existing code), and which alternatives to use instead (ask for Fable, ask_m3, ask_glm, ask_council). It also lists refusal categories, so an agent can avoid out-of-scope asks.

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