friction_digest

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

Annotations already declare read-only, idempotent, non-destructive, closed-world behavior, yet the description adds genuinely new traits: the output is a fixed snapshot, counts are de-duplicated per distinct account/integration, and nothing is logged or submitted. The only missing behavioral note is what the fixed snapshot's recency/window is.

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 short sentences, zero filler, with the core action stated first and the constraints/non-goals following. Every clause ('fixed snapshot', 'counts distinct accounts and integrations separately', 'does not log or submit feedback') carries distinct decision-relevant information.

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?

An output schema exists, so return-value shape needn't be explained, and the description adequately covers scope, determinism, non-goals, and side-effect absence for a simple read-only aggregate. The one real hole is the unexplained 'include_live' parameter, which is enough to keep it short of complete.

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 never mentions the single parameter 'include_live', nor what toggling it does (e.g. whether it mixes live data into an otherwise fixed snapshot). A one-parameter tool with an undocumented, non-obvious boolean leaves the agent guessing at invocation, so the description fails to compensate 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?

The description gives a concrete verb and object: 'Aggregate recurring checklist gaps and failures into Product review questions.' It also implies scope ('counts distinct accounts and integrations separately'), so an agent can tell it is an analytical summary rather than a listing or triage tool. It never names a sibling tool, so differentiation is inferential rather than explicit.

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?

Negative routing exists ('Not a trend, defect diagnosis or model-generated clustering', 'Does not log or submit feedback'), which rules out several plausible misuses and implies this is for snapshot-style review question generation. However, it never says when to reach for this over triage_issue, draft_ticket, or get_account_context, so usage is only implied.

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