prepare_handoff

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

With annotations already declaring readOnly=false and destructive=false, the description adds substantive context beyond them: the server validates and writes files, final_updates are persisted atomically before assembly, and the agent is told what to report back (path, size, token estimate, coverage gaps/warnings). It stops short of saying whether repeated calls create new files or overwrite (relevant given idempotentHint=false) or where files land.

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 trigger condition, then ordering, then server behavior, then the reply contract, then modes — a logical and scannable sequence with no filler. Slightly dense in the final two sentences, where the reply-format instruction reads more like process guidance than tool definition, but each sentence is actionable.

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?

Output schema exists, so return values needn't be documented, yet the description still tells the agent what to surface from the result (primary file path, size, token estimate, 3-line summary, coverage gaps), which is genuinely additive for a complex multi-step tool. The gap is the upload/profile and format options, which never get explained anywhere the agent can see.

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 60% schema coverage, the description carries real weight: it explains the mode enum's meaning and cost (~3k / ~10k default / ~30k / multi-file knowledge_package) and pins down what final_updates must contain (summary, where_stopped, next_action). It leaves 'profile', 'format' and the required 'project' unaddressed, so it is valuable but not complete.

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 artifact ("Prepare a handoff package") plus the exact triggering situations: switching accounts/apps/models or pausing the project. Names the sibling tools it depends on (remember / update_state), so an agent can distinguish it from checkpoint or get_context 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 preconditions and ordering ("FIRST save anything... then call this") and an explicit exclusion ("no re-summarizing" — the server pulls from stored knowledge). It also names the alternatives and what they should be used for first, which is exactly the when-vs-which-tool guidance most definitions lack.

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