pkg

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

With no annotations at all, the description carries the load, and it does add real behavioral context: info is a composite of Libraries.io, Sonatype and PkgSeer (implying federated upstream calls), and upgrade_review bundles vulns + changelog + dependency diff. It does not disclose auth needs, rate limits, upstream failures, or whether operations are read-only, which matters for a 10-param network-backed tool.

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?

One dense, front-loaded sentence packs the action inventory, then a routing rule to the alternative, then a minimal example — no filler. Structure matches how an agent reads: what it is, when to use something else, how to call it.

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 five-action, ten-parameter tool with no output schema and no annotations, the description never indicates what each action returns (fields in a composite info result, changelog entry shape, dependency diff format) or how failures surface. Action coverage is complete, but return-value and error context are absent, which is a meaningful gap given the multiplexing.

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 70%, and the description adds the action-to-parameter mapping the schema does not state as a whole: files/read use path, changelog uses version ranges and count, upgrade_review uses current/target version. The example ('pkg(name=..., action=...)' with the default info) also clarifies the minimal call. Gaps remain on registry='auto' semantics and on version defaults per action.

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?

It names a specific resource (package intelligence) and enumerates all five supported operations (info, changelog, upgrade_review, files, read) with parenthetical scope for each, so the agent knows exactly what the tool can do without opening the schema. It also explicitly distinguishes itself from the search_package sibling for raw single-registry queries.

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 carves a clean boundary against search_package ('for raw single-source registry queries ... use search_package instead') and gives a concrete call example. It stops short of explicit selection rules among the five actions — e.g. when upgrade_review beats changelog + info — leaving that to inference from the parentheticals.

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