Read a visible AGY transcript

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

Annotations include idempotentHint=true, readOnlyHint=false, destructiveHint=false, so the description adds no contradiction. The description does add behavioral context: 'Refresh and return' indicates the operation may re-fetch, and it explicitly states what will not be returned ('Internal thinking, system messages, and checkpoints are never returned.') However, beyond these hints, no extra behavioral traits (e.g., side effects when refresh is true, auth needs, rate limits) are disclosed. It is consistent with annotations; also readOnlyHint=false and idempotentHint=true are fine because a'read' might imply idempotency and no destructive side effect.

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 long, focused, and front-loaded with the core action: 'Refresh and return...'. Every sentence earns its place: the first states scope/object, the second clarifying limitations. No filler or redundant reinforcement of the title.

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 there's no output schema and 5 parameters, the description should explain what the tool returns and any effects of refresh. It states what is returned (user/assistant/tool transcript) and what is excluded, which is good. Still, it doesn't explain the meaning or behavior of the parameters (particularly refresh and conversation_id) and does not describe the return format (e.g., array of messages?). Missing the explicit need for registered conversation context (conversation_id optional?) and an example or caveat around max_chars/max_records. Thus it's adequate but with notable gaps.

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 shows 5 parameters with descriptions for project_root only (coverage 20%). The description mentions 'project-local' and 'bounded', giving light semantic context to project_root and max_chars/max_records. However, the description does not clarify conversation_id, refresh, max_chars, max_records beyond the schema's default and maximum. It adds a small amount of meaning (e.g., 'bounded' suggests records are limited, 'project-local' ties to project_root), but leaves all other parameter semantics up to the schema, so a 3 is reasonable for partial compensation.

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 title 'Read a visible AGY transcript' already clearly indicates a read operation, and the description enhances it with specificity: 'Refresh and return the bounded project-local user/assistant/tool transcript for a registered AGY conversation.' It clearly distinguishes from sibling tools like antigravity_ask or antigravity_get_active_session, though it doesn't explicitly name them as alternatives. The 'bounded' scope and exclusion of internal messages add clarity beyond the title.

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 implicitly states when to use this tool: when you need the user/assistant/tool transcript of a registered AGY conversation. It also provides which exclusions (no internal thinking/system messages/checkpoints), which serves as guidance. However, there are no explicit when-not-to-use alternatives or references to sibling tools like antigravity_sync_conversation or antigravity_get_active_session, which limits clear differentation for the read-only transcript vs. other operations.

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