Skip to main content
Glama

Scenario corpus

corpus

Search, retrieve, list, and register scenario or model fragments by date, region, factors, and kind; index and fetch manifest files as jobs.

Instructions

The scenario corpus and the model corpus. Actions: search, get, list, add, index, fetch.

corpus="model" works with the corpus of the models — the material a model's elicitation reads: episodes, the state at the date, transmission reviews, market expectations — which has its own manifest and index under corpus-model/. search and get read the index as it was on as_of; list shows what the manifest holds; add registers a file downloaded by hand under its key. index and fetch are tasks on the corpus resource. Inside a bench session every call and its answer go to the journal of the event.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "action": "<action>"
    },
    "name": "corpus"
  }
}

Input Schema

NameRequiredDescriptionDefault
kNoHow many fragments search returns, best first.
idNoThe fragment id that get returns, as search reported it.
idsNoThe documents fetch downloads, by manifest key; every missing one when omitted.
keyNoThe manifest key add registers the file under.
urlNoThe address the file was downloaded from, which add records.
fileNoThe path of the downloaded file add registers.
kindNoThe kind of fragment search limits itself to: methodology, scenario_set, transmission, episode or market (market expectations, the model corpus alone).
as_ofNoThe date search and get work as of, YYYY-MM-DD: only fragments visible on that date.
queryNoThe words search looks for in the fragments of the corpus.
actionYesWhat to do: search the index, get one fragment, list the manifest, add a file, or index and fetch as jobs.
corpusNoWhich corpus to work with: "scenario", the default, or "model" — the corpus of the models, with its own manifest and index.scenario
regionNoThe region search limits itself to.
factorsNoThe factors search limits itself to.

Output Schema

NameRequiredDescriptionDefault

No arguments

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed13 schema fields changedv1.0.2
    • addedInput schema / properties / action / description
      Added value: +"What to do: search the index, get one fragment, list the manifest, add a file, or index and fetch as jobs."
    • addedInput schema / properties / as_of / description
      Added value: +"The date search and get work as of, YYYY-MM-DD: only fragments visible on that date."
    • addedInput schema / properties / corpus
      Added value: +{
      +  "default": "scenario",
      +  "description": "Which corpus to work with: \"scenario\", the default, or \"model\" — the corpus of the models, with its own manifest and index.",
      +  "type": "string"
      +}
    • addedInput schema / properties / factors / description
      Added value: +"The factors search limits itself to."
    • addedInput schema / properties / file / description
      Added value: +"The path of the downloaded file add registers."
    • addedInput schema / properties / id / description
      Added value: +"The fragment id that get returns, as search reported it."
    • addedInput schema / properties / ids / description
      Added value: +"The documents fetch downloads, by manifest key; every missing one when omitted."
    • addedInput schema / properties / k / description
      Added value: +"How many fragments search returns, best first."
    • addedInput schema / properties / key / description
      Added value: +"The manifest key add registers the file under."
    • addedInput schema / properties / kind / description
      Added value: +"The kind of fragment search limits itself to: methodology, scenario_set, transmission, episode or market (market expectations, the model corpus alone)."
    • addedInput schema / properties / query / description
      Added value: +"The words search looks for in the fragments of the corpus."
    • addedInput schema / properties / region / description
      Added value: +"The region search limits itself to."
    • addedInput schema / properties / url / description
      Added value: +"The address the file was downloaded from, which add records."
  2. First observedv0.1.0

TDQS

B3.3/5.0
Behavior3/5

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

Annotations declare readOnlyHint=false, openWorldHint=true and non-idempotent behavior, so the safety profile is already carried structurally. The description adds genuine context beyond that — 'Inside a bench session every call and its answer go to the journal of the event' — plus that 'add' writes a registration and 'fetch' downloads missing documents, but it never says whether 'add' or 'index' overwrite existing entries or what permissions are needed.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The action list is front-loaded, which is good, but the prose that follows is dense and unevenly weighted: three lines are spent enumerating what the model corpus contains (episodes, state at the date, transmission reviews, market expectations) while the default scenario corpus gets no equivalent treatment. Formatting artifacts like doubled backticks around corpus="model" add noise.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

An output schema exists, so return values need not be explained, and the write-vs-read split is partially covered. However, for a 13-parameter, six-action tool with destructive potential via add/index/fetch, the description leaves the default scenario corpus's manifest structure, overwrite behavior, and action-specific preconditions unstated.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so every one of the 13 parameters is already documented, setting the baseline at 3. The description adds some cross-parameter meaning — that corpus="model" selects a separate manifest and index under corpus-model/, and that as_of scopes both search and get — but adds nothing for kind, region, factors, ids, url, or file.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description lists concrete verbs (search, get, list, add, index, fetch) against two named resources, the scenario corpus and the model corpus, so the tool's role is legible. It weakens only in differentiating itself from the sibling tools 'scenario' and 'model', which by name appear to overlap with the two corpora handled here.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Per-action roles are sketched: 'search' and 'get' read the index as of as_of, 'list' shows the manifest, 'add' registers a hand-downloaded file, and 'index'/'fetch' are described as 'tasks on the corpus resource' — a vague phrase. There is no guidance on when to choose the default scenario corpus versus corpus="model", nor when to use this tool instead of the similarly named siblings.

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