spranab/brainstorm-mcp

• adapterstring

  cli only: built-in adapter — claude, codex, gemini, qwen, kimi, cursor-agent, opencode, droid — or 'custom'. Defaults to the provider name.

• apiKeyEnvVarstring

  api only: environment variable holding the API key. Use 'NONE' if no key required.

• argsstring[]

  cli only ('custom' adapter): argv template. Placeholders: {{model}}, {{system}}, {{prompt}}, {{outfile}}.

• backendstring

  cli only: run the Claude CLI against another vendor's coding plan — moonshot, minimax, glm. Overrides 'adapter'.

• baseURLstring

  api only: base URL, e.g. 'http://localhost:11434/v1' for Ollama

• commandstring

  cli only: executable to run. Required for adapter 'custom'.

• defaultModelstringобязательный

  Default model for this provider, e.g. 'llama3', 'sonnet'. Use 'default' to let a CLI pick its own.

• kindenum

  'api' (default) for an OpenAI-compatible HTTP endpoint, 'cli' for a local agent CLI.

• namestringобязательный

  Provider name, e.g. 'groq', 'ollama', 'mistral', 'claude'

• promptViaenum

  cli only ('custom' adapter): how the prompt reaches the CLI.

• contextstring

  Optional context to ground the debate — code snippets, PR diffs, error logs, architecture docs, etc. Models will see this alongside the topic.

• modeenum

  Execution mode. Must be provided — if not, the tool will return options for the user to choose. 'api' — MCP server calls model APIs directly using configured API keys. 'hosted' — No API keys needed. Returns prompts for the HOST to execute using sub-agents with models available in the environment (opus/sonnet/haiku, GPT, Gemini, etc.).

• modelsstring[]

  Optional: specific models to use as 'provider:model' (e.g. 'openai:gpt-4o', or 'claude:sonnet' / 'codex:default' for subscription-backed CLI providers). If not provided, all configured providers are used with their default models.

• participateboolean

  Whether Claude should actively participate as a debater in each round (default: true). Set to false for a non-interactive debate between external models only.

• roundsinteger

  Number of debate rounds (default: 3)

• styleenum

  Debate style. 'freeform' (default): open brainstorming. 'redteam': adversarial — models find flaws, risks, and weaknesses. 'socratic': probing questions that expose assumptions and push for deeper understanding.

• synthesizerstring

  Optional: model for final synthesis as 'provider:model'. Defaults to the first model.

• systemPromptstring

  Optional system prompt to guide the brainstorming style or constraints

• topicstringобязательный

  The topic, question, or prompt to brainstorm about

• responsesobject[]обязательный

  Array of model responses collected by the host

• session_idstringобязательный

  The session ID from the brainstorm tool

• contextstring

  Optional context — code snippets, error logs, etc. Models will see this alongside the topic.

• modelsstring[]

  Optional: specific models as 'provider:model' (API providers, or CLI providers like 'claude:sonnet' that use a subscription instead of API credits). If not provided, all configured providers are used.

• styleenum

  Perspective style (default: freeform)

• topicstringобязательный

  The question or topic to get quick perspectives on

• responsestringобязательный

  Your substantive contribution to this round of the debate. Engage deeply with the other models' responses. Minimum 50 characters.

• session_idstringобязательный

  The session ID returned by the brainstorm tool

• descriptionstring

  Optional: PR description or commit message

• diffstringобязательный

  The unified diff to review (e.g., output of git diff)

• focusenum[] | enum

  Optional: focus areas for the review. Default: all categories.

• instructionsstring

  Optional: repo-specific review instructions (e.g., 'we use strict null checks', 'focus on SQL injection risks')

• modelsstring[]

  Optional: specific models as 'provider:model'. Default: all configured providers.

• titlestring

  Optional: PR title or change summary for context

Без параметров.