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

The readOnly/openWorld/destructive annotations already establish this as a safe local read. The description adds genuinely non-obvious behavior beyond them: the read is incremental (only new output since the previous read) and the response includes liveness state. It stops short of covering error cases (stale/finished pid) or truncation behavior.

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 sentence, front-loaded with the primary effect and appending the secondary return value. No filler, nothing repeated from the schema.

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?

With no output schema, the description usefully states what comes back (new output + running status), and wait_ms documents the blocking model. Minor gaps: what happens for an unknown/finished pid, and whether returned output is truncated or bounded.

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 100% — pid's provenance, device's optionality, and wait_ms's default/max are all documented in the schema. The description adds no parameter-level detail, so baseline 3 applies.

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?

States a specific verb and resource ('Return what a background process has printed') plus the key scoping qualifier 'since you last read it', which is exactly what separates it from a full-output read or from interact_with_process. An agent can tell instantly what it gets back: incremental stdout/stderr plus a running flag.

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?

Usage is only implied: the pid schema note ties it to start_process, and 'since you last read it' hints at the polling pattern. There is no explicit statement of when to choose this over interact_with_process, list_processes, or a final read after the process exits.

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