selftest

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

Annotations declare non-read-only, non-idempotent, non-destructive, and the description is consistent: it discloses that the full path actually sends a real test message to a peer agent and blocks waiting for a correlated reply, with an explicit wait window (60-600 s). That is meaningful behavior beyond the annotations. It stops short of saying whether the test message persists in the recipient's inbox or what happens on timeout, which keeps it from a 5.

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 behavior is stated first and the argument notes follow, with essentially no filler. The closing sentence about returning 'the table of both directions' is partly redundant given an output schema exists, and the landmark phrasing is jargon-heavy, but the whole thing is short and earns most of its 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?

For a two-parameter tool with an output schema, the description covers what it does, both modes, both parameters, and the wait bound. The remaining gap is operational: no statement of prerequisites (are the peer agents required to be running?) or what failure/timeout looks like, which matters for a tool that can block for up to ten minutes.

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%, so the description must carry the load, and it largely does: `to` is documented with its two valid values ('codex' or 'antigravity') plus the omitted-means-landmarks-only semantics, and `timeout_s` is documented as the maximum wait with its 60-600 s range. It omits the default value (420) that the schema supplies and does not say what an invalid recipient does.

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 specific verb-and-resource picture: it checks interface landmarks of Codex/Antigravity and, optionally, performs a real end-to-end wake (send test message, wake, read, correlated reply). This is clearly a diagnostic/self-test tool rather than any of the messaging siblings (send, reply, wait_reply). It does not explicitly contrast itself with nearby diagnostic siblings like status, where, or survey, which keeps it at 4 rather than 5.

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 two operating modes are spelled out cleanly: 'with `to`, a real end-to-end wake' versus 'omitted = landmarks only', which tells an agent exactly which mode a given call will trigger. What is missing is any statement of when to prefer this over siblings such as status or where, or a warning about the cost of the full path.

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