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

Annotations already declare readOnlyHint=false but destructiveHint=false, which is confusing on its face; the description resolves that tension by explicitly stating the fragment is staged and never mutates the model, which is real added context. It also discloses 'Free' (no credit cost), a trait absent from every structured field, and implies the provenance requirement. It does not mention whether re-submitting the same fragment is idempotent, which annotations mark false.

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?

Four short sentences with no filler; the cost signal and the staging guarantee are front-loaded ahead of the next-step instruction. The terse 'Free.' fragment is slightly abrupt but informative rather than wasteful.

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 high-complexity tool with a deeply nested schema and no output schema, the description covers cost, staging semantics and next steps well. It omits what the call returns (a staged fragment id?), how branch interacts with the eventual merge, and how id_hints affect identity resolution — gaps an agent would notice on first use.

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 67%, so the schema already documents model_id, subsystem, branch, parent and most provenance fields in detail. The description adds the useful high-level split of payload into 'SysML v2 text or element JSON' and calls out that provenance must accompany the fragment, but says nothing about branch targeting, id_hints, or the parent field — the parts most likely to be misused. Adequate, not compensating for the coverage gap.

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 and resource ('Stage a generated subsystem') and names the accepted payload forms (SysML v2 text or element JSON). It also distinguishes itself from the staging/commit siblings by stating it never changes the model and by pointing to mbe_reconcile as the follow-on step, so an agent can tell it apart from mbe_commit and mbe_reconcile.

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?

'Staged only: it never changes the model. Then call mbe_reconcile.' gives a clear workflow contract and implicitly rules out using this tool to persist changes. It does not explicitly name the alternative for direct commits (mbe_commit) or state preconditions such as needing a baseline from mbe_get_context, so it stops short of the when-not/alternatives bar.

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