Ask Drex decision models typed questions about a state and get calibrated probabilities; free within per-network limits.
Scored across 2 tools
The two tools serve clearly distinct purposes: `decide` submits a prediction query, while `list_models` enumerates available models. There is no plausible confusion between querying probabilities and discovering model options.
Both names are lowercase and action-oriented, but the convention is slightly mixed: `decide` is a bare verb while `list_models` follows a verb_noun pattern. This is a minor deviation but still easily readable.
Two tools is thin for a prediction service, though the surface is narrowly scoped. The core operation and a discovery helper are present, but it sits at the low end of usefulness.
The essential workflow of listing models and requesting calibrated probabilities is covered, but there is no way to inspect a model in detail, retrieve past decisions, or manage API key state. These notable gaps limit agent autonomy.
Ask Drex typed questions about a state and get calibrated probabilities back. Free without an API key, up to 10 calls a minute and 100 a day per network. When the limit is reached the tool says how to add a Drex API key, which runs calls on the key's account at its own limits.
| Name | Required | Description | Default |
|---|---|---|---|
| model | No | The model id. Defaults to drex-latest. A pinned id never changes under you; list_models lists them. | |
| state | No | What every question is about: text, or a JSON object or array, sent as it is. Text only: images, audio, video and PDFs are refused. | |
| questions | Yes | Question name to question, up to 512. The name keys the answer, so name it for what it decides (wants_refund, not q1). noul returns the probability of yes, choice returns one of its labels with a probability for each, score returns the expected level. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations at all, the description carries the behavioral load and does well: it discloses the unauthenticated rate limits (10/min, 100/day per network), that exceeding them surfaces API-key instructions, and that keyed calls run on the key's own quota. It does not describe the shape of the returned answers or latency, which keeps it from a 5.
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?
Three front-loaded sentences with no filler; the capability comes first and the quota/fallback mechanics follow. Slight redundancy in restating the API-key escalation twice, but every sentence carries operational value.
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 tool with a deeply nested schema, no output schema, and no annotations, the description covers auth and limits but says nothing about what a response looks like (per-question probabilities vs. labeled choices vs. expected level) or how 512-question batches behave. Adequate, not 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 100%, so the schema already documents model, state, and the question-type union in detail. The description's 'typed questions about a state' echoes rather than extends that, so the baseline 3 is appropriate.
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 concrete verb and resource: ask typed questions about a state and receive calibrated probabilities. It is specific enough to distinguish from the only sibling (list_models, which enumerates model ids), but the term 'Drex' and 'typed questions' are left for the schema to define, so the purpose is clear rather than crisp.
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?
Explains the free tier and quota escalation, but never says when to reach for this tool versus alternatives or what kinds of tasks it suits (classification, scoring, boolean decisions). No exclusions or prerequisites are given beyond the media-type restriction which lives in the schema.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List the models decide can use. Free, and needs no API key.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full behavioral disclosure burden. It usefully adds that the tool is free and needs no API key, but it does not describe return format, rate limits, or other operational details.
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 two short sentences with no wasted text, and the core purpose is front-loaded. It is appropriately sized for a simple no-parameter listing tool.
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 simple, no-parameter, no-output-schema listing tool, the description gives enough context to call it correctly, including its cost and authentication profile. It could be slightly more complete by mentioning what the returned model list looks like.
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?
The tool has zero parameters, and the schema is empty with full description coverage by definition. The description appropriately does not need to explain parameter semantics, matching the baseline for no-parameter tools.
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?
The description clearly states a specific verb and resource: "List the models" that "decide can use." It is understandable without opening the schema, though it does not explicitly differentiate this capability from the sibling tool "decide."
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 description implies this is a discovery tool for seeing which models are available, and notes that it is free and requires no API key. However, it does not explicitly say when to use this tool versus the sibling "decide" tool or when not to use it.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
decidelist_models
Add one secure layer between your agents and this server.