counsel

Make confident decisions with multi-perspective analysis, debate, and research.

CommunityOnlineNo sign-inGlobalFreeRead-only

What it can do

  • Start Consultation: Start a new strategic consultation, debate, or research task. This is the PRIMARY entry point for any analysis. Use this tool to: - Get a quick second opinion ('quick' mode). - Run
  • Get Consultation Status: Check the progress/status of a running analysis. USE THIS WHEN: - You called start_consultation and need to check if it's done - User asks "is my analysis ready?" or "what's t
  • Get Consultation Report: Retrieve the final results, decision card, and artifacts. USE THIS WHEN: - get_consultation_status shows status='completed' - User asks "what's the result?" or "show me the an

What data it sees

Do you need an account

No: the server works without sign-in

Make confident decisions with multi-perspective analysis, debate, and research. Refine questions, verify claims, and gather evidence to de-risk choices. Personalize guidance and receive clear summaries, recommendations, and trade-off insights.

How to connect

How to connect counsel to Claude

The server address to paste into Claude settings:

https://counsel--mercurialsolo.run.tools

  1. Open Claude (claude.ai or the desktop app).
  2. Go to Settings → Connectors.
  3. Click Add custom connector.
  4. Paste the server address copied below into Remote MCP server URL and click Add.
  5. In a chat, click + → Connectors and switch the new connector on.

Custom connectors are available on Free, Pro, Max, Team and Enterprise plans (Free is limited to one). On Team and Enterprise an organization Owner adds the connector first under Organization settings → Connectors.

Authorization

After you click Add, a sign-in window for the service opens. Sign in with your own account and approve access. Claude never sees your password.

How to connect counsel to ChatGPT

The server address to paste into ChatGPT settings:

https://counsel--mercurialsolo.run.tools

  1. Open ChatGPT in a browser (chatgpt.com). A Plus, Pro, Business, Enterprise or Edu plan is required.
  2. Turn on developer mode once: Settings → Apps → Advanced settings → Developer mode.
  3. Open Settings → Connectors and click Create.
  4. Fill in the form: Name (anything), Description (one line about what the service does), MCP server URL (copy it below).
  5. Under Authentication choose OAuth if the service requires sign-in, otherwise None. Click Create.
  6. In a new chat open + → Apps/Connectors and enable the connector.

OpenAI has renamed this section before (Connectors → Apps/Plugins). If the label differs, search settings for "developer mode". On Business/Enterprise workspaces an admin must allow custom connectors first.

Authorization

On first use ChatGPT opens the service's sign-in window. Sign in and approve access.

How to connect counsel to Cursor

The server address to paste into Cursor settings:

https://counsel--mercurialsolo.run.tools

Fastest: click Open in Cursor below and confirm the prompt.

Manually:

  1. In Cursor open Settings → Cursor Settings → MCP and click Add new global MCP server.
  2. Paste the JSON copied below into ~/.cursor/mcp.json (per project: .cursor/mcp.json in the repo root).
  3. Save the file. The server appears in the MCP list; authorize it there if asked.

Authorization

If the service needs sign-in, an authorize button appears next to the server in the MCP list.

Server tool list (12)

Raw names from tools/list. Only developers need these.

start_consultationStart a new strategic consultation, debate, or research task. This is the PRIMARY entry point for any analysis. Use this tool to: - Get a quick second opinion ('quick' mode). - Run a full multi-perspective debate ('standard' mode). - Conduct deep research and analysis ('deep' mode). - Gather evidence on a topic ('research' mode). USE THIS WHEN: - User asks for strategic advice, analysis, or a second opinion - User wants pros/cons on a decision or idea - User says "debate this", "analyze this", "what do you think about..." - User needs help evaluating options or trade-offs - User wants research on a topic before making a decision DO NOT USE FOR: - Simple factual questions (use your own knowledge) - Non-strategic queries like definitions or explanations - Tasks that don't involve decision-making or evaluation Args: question: The strategic question, topic, or claim to analyze. mode: The depth/type of analysis: - 'quick': fast 30s pros/cons analysis (use for "quick take", "quick check", or very simple questions). - 'standard': rigorous debate and detailed analysis (use for "deliberate", "debate", "consult"). This is the DEFAULT for most strategic questions. - 'deep': full pipeline with web research (use for "analyze", "research project", "deep dive", or when external information is needed). - 'research': only gather evidence without debate (use for "find papers", "search web"). sources: Optional list of sources (documents, context) to use. If omitted, internal knowledge and context will be used. context: Background context or situation description to inform the analysis. analysis_id: Optional custom ID for tracking. counsel_id: Optional ID of a specific counsel/persona configuration to use.
get_consultation_statusCheck the progress/status of a running analysis. USE THIS WHEN: - You called start_consultation and need to check if it's done - User asks "is my analysis ready?" or "what's the status?" - Polling for completion before calling get_consultation_report - Checking if a running analysis needs user input (pauses/steering) WORKFLOW: Call this after start_consultation, repeat until status='completed', then call get_consultation_report for final results. Args: analysis_id: The ID returned by `start_consultation`. include_intervention_status: If True, checks if user input is needed (pauses/steering). include_transcript: If True, includes full debate transcript (slower). Use False (default) for fast polling, True when user wants progress details. Returns: Dict with 'status' (running/completed/failed), 'phase', and 'progress'.
get_consultation_reportRetrieve the final results, decision card, and artifacts. USE THIS WHEN: - get_consultation_status shows status='completed' - User asks "what's the result?" or "show me the analysis" - You need to present the final recommendation to the user - Fetching the decision card, pros/cons, or executive summary DO NOT USE IF: - Analysis is still running (check status first) - Analysis failed (status='failed' - handle error instead) Args: analysis_id: The ID of the completed analysis. include_tensions: Whether to include the 'Crux Registry' (key tensions/trade-offs). presentation_format: Format to render results (e.g., 'brief', 'detailed', 'memo'). ctx: Request context to read client hints (e.g. locale). Returns: The final analysis output, including recommendation, rationale, and key factors.
manage_consultationIntervene in or modify a running analysis. USE THIS WHEN: - User wants to pause, stop, or cancel a running analysis - User says "focus more on X" or "consider Y" mid-analysis (use 'steer') - User wants to add new evidence during an ongoing debate - User wants to edit the identified trade-offs or tensions DO NOT USE FOR: - Starting a new analysis (use start_consultation) - Checking status (use get_consultation_status) - Getting results (use get_consultation_report) Args: analysis_id: The analysis to manage. action: The specific intervention to perform. payload: Data required for the action: - for 'steer': string guidance ("Focus more on X"). - for 'add_sources': list of source objects. - for 'pause'/'abort': optional reason string. Returns: Success status of the intervention.
list_resourcesList available resources (past analyses, configs, sessions). USE THIS WHEN: - User asks "show my past analyses" or "list my debates" - User wants to see available counsel configurations - User needs to find a previous analysis to reference - Browsing available templates or persona configurations Args: type: What to list: - 'consultations': Past debates/analyses. - 'counsels': Available persona configurations. - 'advisor_sessions': Active advisor chat sessions. - 'templates': Pre-built counsel templates. limit: Max items to return. offset: Pagination start index. Returns: List of resource items.
sharpen_questionRefine and clarify a strategic question BEFORE analysis. USE THIS WHEN: - User's question is vague, broad, or unclear - Question could mean multiple things - You want to ensure high-quality analysis by improving the question first - User says "help me frame this" or "what's the right question to ask?" DO NOT USE FOR: - Questions that are already specific and well-defined - When user explicitly says "analyze this exact question" - Simple queries that don't need refinement WORKFLOW: Call this before start_consultation if the question seems unclear. Use the refined question in start_consultation for better results. Args: question: The initial (potentially vague) question. context: Context about the user's situation. auto_select: If True, automatically pick the best refinement (good for automation). Returns: 'selected_question' (the refined version) and 'enhanced_questions' (options).
verify_claimFact-check a specific statement or claim. USE THIS WHEN: - User asks "is this true?" or "verify this claim" - Checking specific facts or statistics for accuracy - Validating assumptions before making decisions - User wants to fact-check something from an analysis DO NOT USE FOR: - Strategic analysis (use start_consultation instead) - Opinions or subjective statements that can't be fact-checked - General research questions (use start_consultation with mode='research') Args: claim: The specific statement/sentence to verify. context: The surrounding text/context where the claim appears. user_sources: Optional list of sources to check against. Returns: Verdict (Supported/Refuted/Ambiguous) and evidence trace.
consult_advisorInteract with the AI Advisor for personalized intake/tuning. USE THIS WHEN: - User wants to refine their profile or requirements - User wants the system to learn their preferences - Starting a new intake session Args: message: The user's message or answer to a question. session_id: Existing session ID. If omitted, starts a new session. action: 'chat' to convert/reply, 'finalize' to complete intake. Returns: Session state and Advisor reply.
get_advisor_profileRetrieve the learned user profile from an advisor session. USE THIS WHEN: - User asks "what do you know about me?" - Checking the results of an intake session Args: session_id: The advisor session ID. Returns: The constructed user profile.
get_pricingGet current pricing, token costs, and user balance. USE THIS WHEN: - User asks "how much does this cost?" or "what's the pricing?" - User wants to check their credit balance - User asks about purchasing analysis credits Returns: Pricing details and current balance.
create_checkoutCreate a secure checkout session to purchase credits. USE THIS WHEN: - User explicitly asks to buy credits (e.g., "buy starter pack") - User confirms a purchase Args: sku: The product SKU to purchase. Returns: Checkout session ID and status.
complete_checkoutFinalize an Instant Checkout transaction. This is NOT called by the user. It is called by the ChatGPT host when a user completes a payment in the Instant Checkout UI. Args: checkout_session_id: The session ID we generated. buyer: Information about the buyer. payment_data: Token and payment details from the host. Returns: Confirmation object conforming to ACP spec.