ntulearn_read_file_content

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

Annotations already cover the safety profile (readOnlyHint, idempotentHint, non-destructive), but the description adds substantial behavior the annotations can't convey: the 1 MB MCP payload cap, ~3K vision tokens per page, per-file 25 MB / batch 40 MB / 50-rendered-page caps, and the fact that oversized or unsupported files land under `skipped` rather than failing. That is exactly the operational context an agent needs before invoking.

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 purpose and the sibling distinction before any parameter detail, then it moves to mode guidance, fallbacks, and caps. It is dense and nearly every sentence carries a distinct constraint, but at roughly 200 words it is long for a read tool and includes some redundancy with the schema's own mode/pages descriptions.

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?

An output schema exists, so return-value shape need not be explained, and annotations already carry the safety semantics. The description fills the remaining gaps a caller would hit — supported formats (Office variants with speaker notes/sheets, text-like files with HTML tag stripping), caps, the skipped-file behavior, and where to go when unsupported. Nothing material 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 coverage is 100%, so the baseline is 3, but the description adds decision rationale beyond the schema: it explains *why* text is the default ('cheap and almost always what you want for written content') and when vision is justified, and reinforces the mode+pages pairing. It is largely additive rather than duplicative, though a few points restate the schema.

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?

Opens with a specific verb+resource ('Read the content of files attached to a Blackboard content item') and immediately scopes the behavior ('returned inline (no local-filesystem hop)'). It also names the sibling it is not — 'ntulearn_download_file is for users who actually want the bytes on disk' — so an agent can distinguish the two without opening either 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?

Explicitly states when to use this ('ask questions about lecture material') vs the alternative ('users who actually want the bytes on disk'), then gives mode selection guidance (text default, vision for diagram/equation/screenshot-heavy), page-range pairing to stay under the 1 MB cap, and a fallback path for multi-page diagram decks. Even the unsupported-binary case is routed ('fall back to ntulearn_download_file for those').

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