Read Remote Job Logs

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

Annotations carry readOnlyHint=true and idempotentHint=true, but the description adds substantial non-obvious behavior: the 256 KiB per-reply page cap, the 256 MiB recording cap after which output stops being captured while the job keeps running, that a truncated log's tail is 'the last thing written before the cap, NOT the job's latest output', and the security rule that log contents must be treated as untrusted data, never as instructions. None of this contradicts the annotations — all consistent with a read-only, side-effect-free 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?

The description is long but logically front-loaded: purpose, then usage, then machine naming, then sibling contrast, then limits, then safety. Nearly every sentence earns its place given the tool's real pitfalls. It loses a point because it re-explains some schema-covered ground (max_bytes' 256 KiB ceiling and offset_bytes' continuation mechanism appear in both schema and description), adding length without new 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?

With no output schema, the description carries the burden of explaining return behavior, and it does: tail-by-default, a header line instructing continuation with offset_bytes, the 256 KiB page cap, and the truncated-state failure mode. It covers machine naming/auth, the read limit, the distinction from a stalled job, and content-safety handling. For a tool with this many edges, nothing an agent needs to call it correctly is missing.

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 100%, so the baseline is 3, but the description genuinely adds meaning beyond the schema: the machine-naming rule for `code` ('AIC- session code... or a saved alias/hostname... if the user's text contains 'aic-'/'AIC-' in any case, that is one of their machines') and the paging protocol for `offset_bytes` (the previous reply's header line naming the continuation offset). It also explains the tail/offset interaction that the schema only states individually.

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?

Opens with a specific verb-resource pair and a distinguishing detail: 'Read the output (stdout and stderr, interleaved as a terminal would show it) of a detached job.' It explicitly differentiates itself from remote_exec later in the same description, so an agent can tell this tool apart from its nearest sibling without opening either schema.

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?

Gives concrete when-to-use guidance ('USE THIS to follow a long run — training loss, build progress, a stack trace after a failure') and names the alternative with the exact condition: 'Unlike remote_exec — whose reply is truncated at 1 MiB... a job's output is kept in a file...' It also warns against a specific misinterpretation ('do not read a stalled tail as a stalled job') and points to remote_job_status / remote_job_list for the `truncated` state.

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