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

Goes well past the readOnly/openWorld annotations by disclosing the return modality per MIME type, a 32,000-character text cap, base64 fallback for binaries, a 10-minute link expiry, and a prompt-injection warning that file contents are untrusted data. These are exactly the operational traits an agent cannot infer from structured fields.

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-loads what the tool does, then sequences type-specific behavior, the escape hatch, the link caveat, and the security note. Every sentence carries distinct information with no restatement of the name or annotations.

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 three-parameter read tool with no output schema, the description covers return formats, size limits, link lifetime, and the trust boundary on file contents. Nothing an agent needs in order to call it correctly and interpret the result 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?

With schema coverage at 33%, the description carries the burden and does: it explains when include_content is required versus unnecessary, and ties project_id/file_id to the notion of a private, project-scoped file. It stops short of noting that both identifiers are UUIDs, though the schema already declares that format.

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 and resource ('Get a private file by project and file id') and immediately scopes it to delivery through this connection, which separates it from list_files and upload_file. An agent can tell without opening the schema that this is a per-file retrieval, not a listing.

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?

Gives clear conditional guidance: images and text need no flag, while PDFs and archives require include_content. It also warns that the download link may be unreachable from sandboxes, which tells the agent when to rely on inline content. It never explicitly points to list_files for discovering file ids, which is the one obvious missing routing cue.

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