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

Annotations already declare non-read-only, non-idempotent, non-destructive and closed-world, so the safety profile is covered. The description adds genuinely useful behavior beyond that: edits are applied sequentially against the evolving text, an empty oldStr appends, and a non-matching passage fails with a `validation` error that also returns the current notes. That failure/rollback-ish detail is exactly the kind of 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?

Three sentences, front-loaded with the mechanism before the content policy and the failure mode. The middle sentence is long but every clause (what to record, one line each, what never to record) carries distinct instruction, so nothing is padding.

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?

No output schema exists, and the description covers the one return case an agent needs to anticipate: a failed match yields a `validation` error plus the current notes. For a two-parameter tool with a flat schema, that is nearly complete; the only gap is identifying the target website and how it relates to the other websiteKnowledge tools.

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 50%: `edits` is well documented in the schema (including oldStr exact-match and replaceAll), while `websiteId` has no description anywhere. The description repeats the append semantics of `oldStr` and the ordering of `edits` but adds nothing new about either parameter and says nothing about which website's notes are targeted. Baseline 3 is appropriate given the partial coverage.

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: 'Applies exact-match text edits to the agent notes, in order.' That is far more precise than a generic 'edit', and the append case is spelled out. It does not name or distinguish itself from plausible siblings such as websiteKnowledge_update or websiteKnowledge_project_edit, so the agent must infer the boundary from the tool name alone.

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?

The description gives strong content-level guidance about what belongs in the notes (durable facts, decisions, rejected directions, constraints; never progress logs or anything derivable from source), which implicitly tells the agent when this tool is appropriate. However, it never names an alternative or states a condition that would route the agent to websiteKnowledge_update or websiteKnowledge_project_edit instead.

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