Skip to main content
Glama
FallDownTheSystem

Converse MCP Server

by FallDownTheSystem
OverviewSchemaSourceRelated ServersDiscussionsAdmin
JavaScript
Remote

chat

Ask one or more AI models in parallel, consensus, or roundtable modes to compare answers, refine decisions, or continue multi-turn threads.

Instructions

UNIFIED CHAT — talk to one or more AI models. mode "chat" (default): 1..N models answer independently in parallel. mode "consensus": ≥2 models answer, then refine after seeing each other. mode "roundtable": models answer sequentially, each building on the running transcript. Supports files, images, and continuation_id for multi-turn threads (you may switch modes on resume). Use model "auto" for automatic selection. IMPORTANT: use the "files" parameter to share code/file content instead of pasting into the prompt.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
modeNoExecution mode. "chat" (default): independent parallel answers. "consensus": ≥2 models answer then refine via cross-feedback. "roundtable": sequential turn-based dialogue in the given model order. Default: "chat".
asyncNoExecute in the background. When true, returns a continuation_id immediately and processes the request asynchronously; poll with check_status. Default: false
filesNoFile paths to include as context (absolute or relative). Supports line ranges: file.txt{10:50}, file.txt{100:}. Example: ["./src/utils/auth.js{50:100}", "./config.json"]. IMPORTANT: Always use this parameter to share file content instead of copying code into the prompt.
exportNoExport the conversation to disk. Creates a folder named for the continuation_id with numbered request/response files and metadata. Default: false
imagesNoImage paths for visual context (absolute or relative paths, or base64 data). Example: ["C:\Users\username\diagram.png", "./screenshot.jpg", "data:image/jpeg;base64,/9j/4AAQ..."]
modelsNoModels to use. Examples: ["auto"] (recommended), ["codex"], ["codex", "gemini", "claude"], ["codex:astra"], ["gpt-6-astra"]. Forms: "provider" (its default model), "provider:model" (that provider only), or a bare "model" (served by the first configured provider that offers it, local CLI providers first, failing over to the next). Providers: codex, gemini (agy), claude, copilot, openai, google, xai, anthropic, mistral, deepseek, openrouter, abliteration. Unknown names are rejected with suggestions. In mode "chat" each model answers independently; in "consensus" they refine after seeing each other; in "roundtable" they speak in the given ORDER, each seeing the transcript. Default: ["auto"].
promptYesYour question, topic, or task with relevant context. More detail enables better responses. Example: "How should I structure the authentication module for this Express.js API?"
continuation_idNoContinuation ID for a persistent multi-turn thread. Auto-generated in the first response; pass it back to continue. You MAY change the mode or models on a resuming turn.
reasoning_effortNoReasoning depth for thinking models, weakest to strongest: "none" (reasoning off, where the model allows it), "minimal", "low", "medium" (balanced), "high", "xhigh", "max". Passed through by name when the model accepts it, otherwise clamped to the nearest tier it does. Default: "medium"

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changedv4.5.0
    • changedInput schema / properties / models / description
      Previous value: -"Models to use. Examples: [\"auto\"] (recommended), [\"codex\"], [\"codex\", \"gemini\", \"claude\"], [\"codex:astra\"], [\"gpt-6-astra\"]. Forms: \"provider\" (its default model), \"provider:model\" (that provider only), or a bare \"model\" (served by the first configured provider that offers it, local CLI providers first, failing over to the next). Providers: codex, gemini (agy), claude, copilot, openai, google, xai, anthropic, mistral, deepseek, openrouter. Unknown names are rejected with suggestions. In mode \"chat\" each model answers independently; in \"consensus\" they refine after seeing each other; in \"roundtable\" they speak in the given ORDER, each seeing the transcript. Default: [\"auto\"]."New value: +"Models to use. Examples: [\"auto\"] (recommended), [\"codex\"], [\"codex\", \"gemini\", \"claude\"], [\"codex:astra\"], [\"gpt-6-astra\"]. Forms: \"provider\" (its default model), \"provider:model\" (that provider only), or a bare \"model\" (served by the first configured provider that offers it, local CLI providers first, failing over to the next). Providers: codex, gemini (agy), claude, copilot, openai, google, xai, anthropic, mistral, deepseek, openrouter, abliteration. Unknown names are rejected with suggestions. In mode \"chat\" each model answers independently; in \"consensus\" they refine after seeing each other; in \"roundtable\" they speak in the given ORDER, each seeing the transcript. Default: [\"auto\"]."
  2. Changed1 schema field changedv4.0.0
    • changedInput schema / properties / models / description
      Previous value: -"Models to use. Examples: [\"auto\"] (recommended), [\"codex\"], [\"codex\", \"gemini\", \"claude\"]. In mode \"chat\" each model answers independently; in \"consensus\" they refine after seeing each other; in \"roundtable\" they speak in the given ORDER, each seeing the transcript. Default: [\"auto\"]."New value: +"Models to use. Examples: [\"auto\"] (recommended), [\"codex\"], [\"codex\", \"gemini\", \"claude\"], [\"codex:astra\"], [\"gpt-6-astra\"]. Forms: \"provider\" (its default model), \"provider:model\" (that provider only), or a bare \"model\" (served by the first configured provider that offers it, local CLI providers first, failing over to the next). Providers: codex, gemini (agy), claude, copilot, openai, google, xai, anthropic, mistral, deepseek, openrouter. Unknown names are rejected with suggestions. In mode \"chat\" each model answers independently; in \"consensus\" they refine after seeing each other; in \"roundtable\" they speak in the given ORDER, each seeing the transcript. Default: [\"auto\"]."
  3. Changed2 schema fields changedv3.6.0
    • changedInput schema / properties / reasoning_effort / description
      Previous value: -"Reasoning depth for thinking models. Examples: \"none\" (no reasoning, fastest - GPT-5.1+ only), \"minimal\", \"low\", \"medium\" (balanced), \"high\", \"max\". Default: \"medium\""New value: +"Reasoning depth for thinking models, weakest to strongest: \"none\" (reasoning off, where the model allows it), \"minimal\", \"low\", \"medium\" (balanced), \"high\", \"xhigh\", \"max\". Passed through by name when the model accepts it, otherwise clamped to the nearest tier it does. Default: \"medium\""
    • changedInput schema / properties / reasoning_effort / enum
      Previous value: -[
-  "none",
-  "minimal",
-  "low",
-  "medium",
-  "high",
-  "max"
-]New value: +[
+  "none",
+  "minimal",
+  "low",
+  "medium",
+  "high",
+  "xhigh",
+  "max"
+]
  4. First observedv3.0.1

TDQS

A4.1/5.0
Behavior4/5

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

With no annotations, the description carries the full burden and does well: it explains that async returns a continuation_id immediately for background processing, that export writes numbered files and metadata to disk, and that continuation_id enables multi-turn threads with mode switching on resume. It omits any mention of cost, latency, or per-model failure behavior, which is a real gap for a multi-provider fan-out tool.

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

Conciseness4/5

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

Long but front-loaded: the unified purpose and mode taxonomy come first, then supporting capabilities. The 'IMPORTANT: use the files parameter' instruction is duplicated verbatim in the schema's files description, which is minor redundancy rather than bloat.

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

Completeness4/5

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

For a 9-parameter tool with no annotations and no output schema, the description covers modes, asynchronous flow, file/image context, and thread continuity well. Its main omission is return shape — the agent is not told how per-model answers are presented or how consensus/roundtable output is structured, and with no output schema that must come from prose.

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 the schema already documents all 9 parameters in detail, including mode, async, files, models, and reasoning_effort. The description largely restates that schema content (modes, files-vs-prompt advice, model 'auto') rather than adding syntax or edge-case meaning beyond it, so baseline 3 applies.

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

Purpose5/5

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

States a specific verb and resource ('talk to one or more AI models') and immediately enumerates the three execution modes with their distinct semantics. An agent can distinguish this from siblings cancel_job, check_status, and decide without opening the schema.

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

Usage Guidelines4/5

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

Gives clear routing guidance for mode selection, tells the agent to use the files parameter instead of pasting code, and links async=true to polling via check_status. It never contrasts this tool with the sibling 'decide', which appears to be an adjacent model-consultation tool, so the sibling boundary is left implicit.

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

Deploy Server

Other Tools