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

Annotations only declare the write/idempotent/non-destructive profile; the description adds the substantive facts an agent needs — that unspecified fields are left alone, that agents are unversioned, that the response returns 'previous' old values which can be sent back to undo, and that the user should be told what changed. This is meaningful context beyond the annotation set and compensates for the absent output schema.

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 field list, then short directive sentences ('Pass only what changes', 'Tell the person what changed'). The enumeration is dense but each item is a real parameter; only the max_token parenthetical stretches slightly.

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 9-parameter mutation tool with no output schema, the description covers return value (previous values), undo path, partial-update behavior, and a key token-budget caveat. It does not mention failure conditions such as an invalid agent_id or whether the caller's permission level matters.

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?

At 44% schema coverage the description must carry weight, and it does: it names every editable field and adds real meaning to max_token ('reply size limit, which reasoning models also spend on thinking, so keep it at 4000 or more'), a caveat found nowhere in the schema. It stops short of explaining reasoning_effort, temperature or max_tool_calls 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?

Names a specific verb ('Change') and resource ('an agent') and enumerates exactly which attributes are mutable (name, description, instructions, model, reasoning effort, temperature, max_token, max_tool_calls), implicitly scoping out sibling concerns like tools or knowledge which live in set_agent_tools/set_agent_knowledge.

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?

'Pass only what changes' clearly establishes partial-update semantics, which is the main usage decision for this tool. It does not route the agent away from siblings (set_agent_tools, archive_agent, duplicate_agent) or state any exclusion, so it falls short of a 5.

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