Agent Docs

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

Annotations already declare readOnly, idempotent, non-destructive, and closed-world, so safety is covered; the description adds the genuinely useful behavior of paginated continuation ('continue long sections with next_offset') and what content each mode returns (description + outline vs. full text plus subsections). It stops short of describing result shape details, but an output schema exists to carry those.

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?

Front-loaded with the core purpose, then the two mode-specific behaviors and the one caveat, in roughly five tight lines. Every clause drives behavior; no filler.

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?

With 100% schema coverage, rich annotations, and an output schema present, the description only needs to add mode selection and pagination — and it does exactly that. Nothing an agent needs to call this correctly is missing.

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 description coverage is 100%, so url, heading, offset, and outline_depth are already fully documented in the schema, including the exact/case-insensitive/partial heading match and the 'Parent > Heading' disambiguation. The description restates outline_depth and continuation behavior but adds no syntax or semantics beyond the schema, which is the baseline-3 case.

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 and resource (read a page/section) and immediately anchors it to its only sibling: 'Read a page found by search_docs.' It also names the two operating modes (no heading = outline, with heading = section text), so an agent can distinguish this from search_docs without opening a schema.

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?

Explicit conditional routing: omit heading to get the outline, supply heading to get that section's text and subsections. It also warns against over-using outline_depth=3 ('use it only on long pages to find one specific subsection') and points to next_offset for continuation — when-to-use, when-not-to-use, and the alternative tool are all covered.

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

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

With readOnly/idempotent already covered by annotations, the description adds genuinely new behavioral context: one passage per page, a relevance score with a ~0.5 signal threshold, and the one-topic-per-call constraint. These are operational traits not available in the schema or annotations.

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?

Purpose and scope are front-loaded in the first two sentences, with return-shape and follow-up guidance after. It is denser than strictly necessary (the relevance-score aside could be trimmed), but every sentence carries actionable content.

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 search tool with an output schema, the description covers scope, exclusion criteria, disambiguation, result shape, and the escalation path to read_section. An agent has everything needed to call it and interpret results without opening other fields.

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 coverage is 100%, so the baseline would be 3, but the description adds decision guidance beyond the field docs: set product only when the user has named it (a wrong product hides the right pages), and start concise / re-run detailed with a small limit. This meaningfully shapes parameter usage.

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 a specific verb+resource ('Search the official documentation') and enumerates the exact scope (Claude Code, Cursor, OpenAI Codex, MCP). It also carves itself away from the sibling read_section by naming it as the follow-up for full content.

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?

Explicit when-to-use ('before answering any question about using, configuring or administering these products'), explicit when-not ('Not for other products or general programming'), and an ambiguity rule ('If the answer could differ between products... ask'). It also routes to read_section for deeper content.

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