suggest_next_stepRetrieve the single next experiment step from the reasoning engine.
| Name | Required | Description | Default |
|---|---|---|---|
| bean_id | Yes |
| Name | Required | Description | Default |
|---|---|---|---|
| category | Yes | ||
| warnings | Yes | ||
| next_step | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already establish read-only, idempotent, non-destructive behavior, and the description only adds that the result is a single step from the reasoning engine. It does not add meaningful behavioral context beyond the annotations, but it does not contradict them either.
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, front-loaded sentence with no filler. Every word earns its place, and the description avoids redundantly repeating the title.
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 one-parameter read-only tool with an output schema, the description is nearly adequate for selection, but it leaves parameter semantics and when-to-use context to inference. A brief note on bean_id's role would make it complete.
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% and the description never mentions bean_id or why it is needed. The name 'bean_id' gives minimal self-evident meaning, but the description does not compensate for the missing schema documentation or explain how to select the value for a next-step request.
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?
Uses a specific verb ('Retrieve') and identifies the exact resource ('the single next experiment step') and source ('reasoning engine'). This makes the tool's function unambiguous and distinguishes it from the read-only sibling tools, none of which claim to return a suggested next step.
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 phrase 'next experiment step' implies this is the tool for advancing an experiment, but the description does not explicitly state when to prefer it over other getters like get_rule or get_stats, nor any exclusions or prerequisites. Usage guidance is left to inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Most tools target a clearly distinct resource and action, and the list/register/update/set tool families are easy to tell apart. The closest pair is diagnose_preview and diagnose_shot, which are well-described but similar enough in name that an agent could select the wrong one.
The overwhelming majority of tools follow a consistent verb_noun pattern (list_beans, register_grinder, update_shot, set_active). Minor exceptions like grinder_math and kb_changelog lack the imperative verb prefix, but they are readable and do not create real confusion.
34 tools is above the 25+ threshold and feels heavy even though the domain is fairly rich. The many parallel list_* and register_* tools for beans, grinders, machines, scales, waters, programs, and recipes could plausibly be consolidated or trimmed without losing core capability.
The core shot lifecycle is well covered: log, update, delete, diagnose, and list shots, plus bean registration and maintenance tracking. However, most registered entities lack update/delete tools, and get_rule has no corresponding list_rules tool, leaving some obvious workflow gaps that agents must work around.