althing.dev WebMCP Tools, Schemas & Agent Capabilities

← WebMCP Directory

althing.dev

live Unverified (seed import) AI & Agents 4 tools imperative implementation

Open website How verification works

althing

First seen 2026-11-05 · Last seen 2026-11-05 · Source: webmcp.com

Answer 0 · Act 4 · Transact 0 · Score: Not yet scored

Tool inspector

run_panel

act imperative

Run a full synthetic focus group panel. Each persona answers all questions independently in parallel; a synthesis step then aggregates findings into themes, agreements, disagreements, and recommendations. Supports v1/v2 flat instruments and v3 branching instruments.

Page: /

Input schema

Property	Type	Required	Description
`questions`	array	no	Flat list of question dicts (v1-equivalent). Each should have a 'text' key.
`personas`	array	no	Inline persona definitions. Each persona object accepts 'name' (required), plus optional 'age', 'occupation', 'background', 'personality_traits'.
`pack_id`	string	no	ID of a saved persona pack. Merged with inline personas (inline first).
`instrument`	object	no	Raw instrument body (the value under the top-level 'instrument:' key in YAML). Takes precedence over 'questions'.
`instrument_pack`	string	no	Name of an installed instrument pack. Takes precedence over 'instrument' and 'questions'.
`model`	string	no	LLM model to use. Defaults to haiku.
`response_schema`	object	no	Optional JSON Schema for structured output per panelist response.
`synthesis`	boolean	no	Whether to run synthesis after collecting responses. Defaults to true.
`synthesis_model`	string	no	Model to use for synthesis. Defaults to panelist model.
`synthesis_prompt`	string	no	Custom synthesis prompt. Replaces the default.
`temperature`	number	no	—
`top_p`	number	no	—
`persona_models`	object	no	Per-persona model overrides: maps persona name to model alias.
`extract_schema`	—	no	Schema for post-hoc structured extraction. Built-in name ('sentiment', 'themes', 'rating') or inline JSON Schema.
`models`	array	no	List of model names for multi-model ensemble (length >= 2 enables ensemble mode).
`synthesis_temperature`	number	no	—
`variants`	integer	no	Number of variant runs for sampling-based aggregation.
`use_sampling`	boolean	no	—
`decision_being_informed`	string	no	v1.0.0 contract field — the decision this panel will inform (12-280 chars, single line).

Raw JSON schema

{
  "type": "object",
  "properties": {
    "questions": {
      "type": "array",
      "description": "Flat list of question dicts (v1-equivalent). Each should have a 'text' key.",
      "items": {
        "type": "object"
      }
    },
    "personas": {
      "type": "array",
      "description": "Inline persona definitions. Each persona object accepts 'name' (required), plus optional 'age', 'occupation', 'background', 'personality_traits'.",
      "items": {
        "type": "object"
      }
    },
    "pack_id": {
      "type": "string",
      "description": "ID of a saved persona pack. Merged with inline personas (inline first)."
    },
    "instrument": {
      "type": "object",
      "description": "Raw instrument body (the value under the top-level 'instrument:' key in YAML). Takes precedence over 'questions'."
    },
    "instrument_pack": {
      "type": "string",
      "description": "Name of an installed instrument pack. Takes precedence over 'instrument' and 'questions'."
    },
    "model": {
      "type": "string",
      "description": "LLM model to use. Defaults to haiku."
    },
    "response_schema": {
      "type": "object",
      "description": "Optional JSON Schema for structured output per panelist response."
    },
    "synthesis": {
      "type": "boolean",
      "description": "Whether to run synthesis after collecting responses. Defaults to true."
    },
    "synthesis_model": {
      "type": "string",
      "description": "Model to use for synthesis. Defaults to panelist model."
    },
    "synthesis_prompt": {
      "type": "string",
      "description": "Custom synthesis prompt. Replaces the default."
    },
    "temperature": {
      "type": "number"
    },
    "top_p": {
      "type": "number"
    },
    "persona_models": {
      "type": "object",
      "description": "Per-persona model overrides: maps persona name to model alias."
    },
    "extract_schema": {
      "description": "Schema for post-hoc structured extraction. Built-in name ('sentiment', 'themes', 'rating') or inline JSON Schema."
    },
    "models": {
      "type": "array",
      "description": "List of model names for multi-model ensemble (length >= 2 enables ensemble mode).",
      "items": {
        "type": "string"
      }
    },
    "synthesis_temperature": {
      "type": "number"
    },
    "variants": {
      "type": "integer",
      "description": "Number of variant runs for sampling-based aggregation."
    },
    "use_sampling": {
      "type": "boolean"
    },
    "decision_being_informed": {
      "type": "string",
      "description": "v1.0.0 contract field — the decision this panel will inform (12-280 chars, single line)."
    }
  }
}