split

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

Annotations already establish this is a non-readOnly, non-destructive, non-idempotent mutation, and the description usefully adds what actually changes on disk: index.md created if missing (with a generated list), log.md started, links rewritten, cross-bundle links listed but left intact with a lint warning, and code-block links also rewritten. What it omits is the single most important behavioral fact – that apply defaults to false, so the call likely previews rather than mutates – and the description never says what the returned diff looks like or that repeated runs are unsafe (contradicting nothing, but leaving idempotency unaddressed).

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?

The main action is front-loaded and the content is dense with real information, but the prose is a stack of parenthetical spec citations that interrupt the sentence flow, and the trailing 'shortcut: ... read the diff.' fragment is awkwardly telegraphed and leans on the reader already knowing what diff is being referenced.

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 filesystem-mutating tool with no output schema, three undocumented parameters, and no annotations covering permissions, the description does well on effects but leaves the preview/commit semantics and the `bundle` argument entirely unexplained. An agent could not confidently call this correctly on the first attempt without experimenting.

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 0% and the description compensates for only one of three parameters: `dir` is implied by 'a directory'. Neither `apply` nor `bundle` appears anywhere in the text, so an agent has no idea what `bundle` names (the target bundle id? a name?) or that `apply` toggles the preview/commit mode the description hints at.

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?

The opening clause names a concrete verb and resource – 'Make a directory its own bundle' – and enumerates the resulting effects (index.md with okf_version "0.2", a new log.md, rewritten `/` links and §6.2 paths). It is distinguishable from siblings like init/migrate/rename by describing bundle-root promotion. The spec cross-references (§3, §6.1) are opaque outside the tool's own doc set, and no sibling is named explicitly.

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?

Usage is only implied: you split a directory when it should become an independent bundle. There is no statement of when to prefer split over migrate or init, nor when to set apply versus doing a preview – the 'read the diff' line gestures at a review step but never says the call is a dry run unless applied.

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