Checkpoint

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

Annotations already indicate idempotentHint=true and non-destructive write. The description adds key behavior: 'Effect: draft', server-computed digest, idempotencyKey replay semantics ('never bypasses authorization'), and human approval at a later stage. No contradictions with annotations.

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 description is well-structured with clear section headers (Use when, Do not use when, Requires, Effect, Then, On failure) and front-loads the critical purpose. It is somewhat long, but every sentence adds necessary context; the organization earns a high score for clarity despite length.

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?

Given the high complexity (11 parameters, rich nested objects, output schema present), the description covers purpose, usage conditions, behavioral traits, error handling, and post-invocation steps. It omits detailed parameter semantics but includes key constraints (no credentials, no full conversation). The presence of an output schema partially compensates, but low schema coverage limits completeness.

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?

With schema description coverage at only 36%, the description carries responsibility for explaining parameters. Instead, it provides only high-level terms ('observation bounds', 'safe source spans', 'candidate changes') without clarifying individual fields like metadata, sessionId, evidenceIds, or nested structures like conversationProvenance. The schema's own descriptions are sparse, and the tool label does not compensate.

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 description clearly states the tool creates a 'V1 checkpoint draft' from the visible conversation, using specific verbs and resources. It provides scope ('observation bounds, safe source spans...') and contrasts with 'narrower tool' usage, though sibling tools like commit_checkpoint are not 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?

The description explicitly includes 'Use when' and 'Do not use when' sections with concrete conditions, plus 'Requires', 'Human approval', 'Then' guidance, and a full 'On failure' error-handling table. This leaves no ambiguity about when and how to invoke the tool.

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