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

Annotations only cover the read-only/safe profile; the description goes well beyond them with a size-budget model (~12,000 chars, body_has_more/body_part+1, next_start_page_id), truncation flags, the fact that page ids have gaps so ±1 stepping is wrong, and a named error taxonomy (PAGE_NOT_FOUND, TITLE_NOT_FOUND, BOOK_NOT_FOUND, BOOK_NOT_DOWNLOADED with remediation). This is exactly the operational context annotations cannot convey.

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 overview and the unit list, then rules, then errors, then alternatives — a deliberate and correct ordering. Despite its length, almost every clause carries load (mode behavior, budget, id semantics, error recovery), and the dash-per-unit format is scannable rather than prose-bloated.

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 13-parameter, six-mode dispatcher with an output schema, the description covers everything an agent needs: which unit to pick, which parameters belong to which unit, continuation on truncation, id semantics, and recoverable errors. Return-value shape is delegated appropriately to structuredContent and the output schema.

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 per-parameter unit scoping is largely already in the schema. The description still adds cross-parameter semantics the schema cannot express: 'a parameter of another unit is refused', the continuation contract (feed next_start_page_id back in), and the budget interaction between range and section reads. That is real added meaning, though it does not restate every parameter.

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?

Starts with a specific verb+resource ('Read an installed book's text and structure by internal ids') and then enumerates the six distinct modes, each with what it returns (page → body/foot/printed label/prev-next ids; toc → headings with title_ids; parts → volumes). The closing 'Instead:' block names sibling tools and the conditions that route away from this one, so an agent can distinguish it from shamela_find, shamela_search and shamela_get_details 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?

Gives explicit when-to-use guidance per mode ('to read on past a hit or through a span', 'for a printed «vol./page» call shamela_cite first') plus negative routing: headings by words → shamela_find(kind='heading'), untitled «فصل» headings → shamela_search(match='words'), metadata → shamela_get_details. It also states a hard rule about what will be refused, which is the kind of exclusion guidance that prevents wrong calls.

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