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

Annotations already declare readOnlyHint=false, destructiveHint=false and idempotentHint=true, yet the description adds substantial context beyond them: receipt-state semantics (added/already_present confirmed vs pending reservation vs skipped/failed/conversion-required), the recovery flow with resumeImportId and a new recovery key, and the backend binary capability requirement for business-library PDF/DOCX/XLSX. None of this is present in the 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?

Front-loads the core action and constraints before moving to recovery, receipt interpretation and business-library caveats. Every sentence carries operational information, though the density is high and the business-library sentence is somewhat ancillary to the primary call path.

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 no output schema, the description compensates by explaining what the returned receipts mean (confirmed vs pending reservations vs unconfirmed states) and by covering error recovery, idempotency reuse, and the special conversion-capable document types. An agent has enough to invoke this 8-parameter mutation correctly and interpret outcomes.

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 high (75%), so most parameters are self-documenting, but the description still adds meaning: canonical base64 or UTF-8 text, the ~1 MiB/20-file ceiling, and the prohibition on paths, URLs, guessed content and authority overrides. It also ties resumeImportId to the original-bytes recovery protocol, going beyond the schema text, though it does not fully explain versionOf or retryRejected semantics.

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 ('Transfer real selected source bytes') plus its precondition ('after preview and user authorization'), which cleanly separates it from sibling read/preview tools such as helvabase_preview_sources and upload variants like helvabase_upload_original. An agent can identify the operation and its scope without inspecting the 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?

Gives explicit alternatives and conditions: 'For larger files use bearer-authenticated POST /mcp/files', 'Use helvabase_source_imports after interruption, then resumeImportId', and a negative rule ('Never send paths, URLs, guessed content or authority overrides'). It also flags when a result is not confirmed ingested. This is genuine when/when-not guidance rather than implied usage.

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