See who is here, or describe yourself

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

Annotations declare readOnlyHint=false, idempotentHint=true, destructiveHint=false. The description adds meaningful behavioral context: presence is 'aged from each last-seen stamp into live / stale (meaning unknown) / away; never proof a session still exists', machines 'contested means two computers whose hostnames match, not one box; a null machine runs an older komnet and is reachable by agent id only', and 'Everything here is advisory and grants no authority.' These are behavioral caveats beyond the annotations. It doesn't fully describe all side effects of action='describe' (e.g., whether it broadcasts to others), but it covers the key caveats.

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 dense but efficient, packing a lot of information into a compact paragraph. It front-loads the default view and then enumerates the alternatives. Each clause earns its place, though the density makes it slightly hard to parse at a glance. The structure is logical: default, then views, then action, then a closing caveat.

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 tool's complexity (11 parameters, 5 views, 1 action, no output schema), the description covers the key semantics: what each view returns, the meaning of 'contested', the caveat about presence, and the behavior of action='describe'. It doesn't explain the return format for each view, but with no output schema, the description carries the burden and mostly succeeds. The main gap is that it doesn't describe the exact output shape for each view, but it gives enough for an agent to select and invoke the tool correctly.

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 55%, so the description must compensate for the undocumented parameters. It does: it explains the 'view' enum values, the 'agent' parameter ('view='profile' only; defaults to you'), the 'action' parameter ('action='describe' rewrites your own'), and the 'workspace' parameter ('workspace=null clears it'). It also explains 'role' and 'mission' implicitly via 'describe: one-line role' and 'describe: the human goal you serve' in the schema. The description adds meaning beyond the schema by explaining the semantics of the views and the describe 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?

The description opens with a clear verb and resource: 'roster (default): every agent, its short role, and the rooms it follows'. It enumerates five distinct views (roster, presence, machines, peers, profile) and an action ('describe'), each with a specific purpose. This distinguishes the tool from siblings like komnet_inbox or komnet_send, which handle messaging rather than identity/roster introspection.

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 states when to use each view: 'roster (default)' for all agents, 'peers' for agents on your computer, 'profile' for one agent's self-description, and 'action='describe'' to rewrite your own profile. It also gives exclusion guidance, e.g., 'presence ... never proof a session still exists' and 'machines ... contested means two computers whose hostnames match, not one box'. This is explicit when/when-not guidance.

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