create_job

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

Annotations cover the safety profile (readOnly=false, destructive=false, openWorld=false), so the description's job is added context, and it delivers: what happens when no config is supplied, that `base_job` clones an existing job's config, and that `overrides` mutates config.process[0]. It doesn't say whether creation validates the config, whether the job is written to disk, or what happens on invalid paths.

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?

Front-loaded with the verb+resource, then organized by concern (config source, overrides, gpu, start) rather than as a wall of prose. Dense but every clause carries operational meaning; the only waste is a slightly awkward line break after the first sentence.

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?

This is a multi-parameter mutation tool, and the description covers selection logic, formatting, and the immediate-start flag. Since an output schema exists, return values need not be described, but the absence of any statement about validation, error behavior, or name uniqueness leaves a small gap.

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 carries the full burden and does so well: `config` accepts YAML/JSON of a full file or just the `config:` body, `base_job` takes an id/name, `overrides` uses dotted paths with a concrete worked example, `gpu_ids` takes an index string like "0,1", and `start` queues immediately. Only the obvious `name` is left to the schema title.

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 ("Create a training job in the ai-toolkit UI"), which is unambiguous. It does not explicitly differentiate itself from siblings like `new_job_template` or `start_job` — the default-template mention is the only hint at how it relates to those tools.

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 explicit branching guidance for the mutually exclusive config sources ("pick one: config ... or base_job ...; Neither = the default template") and explains that `start` queues immediately, which implicitly answers when to use this versus a later `start_job`. It stops short of naming sibling alternatives or stating prerequisites/failure modes.

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