submit_job

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

Annotations only declare readOnlyHint=false and destructiveHint=false; the description adds substantial context beyond that – the safe dry-run-by-default flow, the confirm gate to trigger real submission, and the automatic partition/architecture deduction with a warning against overriding. This is exactly the kind of mutation-behavior disclosure annotations cannot carry. No contradiction with the write annotation.

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?

A single dense paragraph that front-loads the core action and dry-run behavior. Every sentence carries information, though the parameter enumeration in the middle is slightly packed and the value would be higher if the most-used params led.

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 25-parameter, write-capable tool the description covers the distinctive knobs and the simulation/confirm lifecycle well, and an output schema exists so return values need not be explained. But the majority of parameters are still undocumented given 0% schema coverage, leaving a meaningful gap for an agent configuring a job.

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 0% across 25 parameters, so the description must compensate, and it partially does by explaining distributed, container, redirect_caches, stage_archive, data_files (with the 20-file/64 MiB cap), confirm, partition, and arch. Roughly two-thirds of the parameters (nodes, mem_gb, modules, time_limit, gpus_per_node, array, secret_env_file, etc.) remain undocumented in both places.

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?

Starts with a specific verb+resource ('Prepare et soumet un job SLURM') and characterizes the distributed families it orchestrates, so the agent knows exactly what it generates. It does not, however, draw the boundary against close siblings like submit_array_job, submit_resilient_job, or submit_pipeline, so differentiation is left to inference.

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 clear usage context: runs in simulation by default (returns the sbatch script, deduced partition/arch, sizing warnings) and requires confirm=true to actually submit. It also says not to force partition/arch unless there is a specific reason. It stops short of naming when to prefer an alternative sibling.

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