Skip to main content
Glama

Humanitarian Negotiation MCP Server

HOW_TO_USE_CHATGPT.md5.79 kB
# How to Use All Tools in ChatGPT - No Copy-Paste Needed! ## The New Way (RECOMMENDED) - Automatic Data Flow Thanks to automatic caching, you **no longer need to copy-paste data** between tools! ### Simple 2-Step Workflow: **Step 1: Analyze Stakeholders** ``` Analyze these stakeholders: Context: [Your negotiation context] Stakeholders: 1. Actor A - Power: 0.8, Urgency: 0.9, Legitimacy: 0.85, Position: 0.6 2. Actor B - Power: 0.7, Urgency: 0.8, Legitimacy: 0.9, Position: -0.7 3. Actor C - Power: 0.5, Urgency: 0.6, Legitimacy: 0.7, Position: 0.5 ``` ChatGPT will return the analysis with all stakeholders and their priority levels. **Step 2: Develop Influence Tactics (FOR EACH STAKEHOLDER)** ``` Now develop influence tactics for [Stakeholder Name] using the latest analysis. ``` That's it! The system automatically retrieves the analysis you just ran. --- ## Complete Example: Riverland Crisis ### Step 1: Run Stakeholder Analysis ``` Analyze these 5 stakeholders involved in the Riverland humanitarian crisis: Context: Complex multi-party humanitarian response requiring coordination between government, armed groups, NGOs, donors, and media. Stakeholders: 1. Riverland Government - Power: 0.9, Urgency: 0.8, Legitimacy: 0.95, Position: 0.6 2. GRN (Armed Group) - Power: 0.7, Urgency: 0.9, Legitimacy: 0.3, Position: -0.8 3. Local NGOs - Power: 0.5, Urgency: 0.8, Legitimacy: 0.7, Position: 0.7 4. Donors - Power: 0.6, Urgency: 0.7, Legitimacy: 0.8, Position: 0.8 5. Media - Power: 0.4, Urgency: 0.6, Legitimacy: 0.6, Position: 0.5 ``` **Response:** Analysis with all 5 stakeholders, their salience scores, and priority levels (First/Second/Third). ### Step 2a: Develop Tactics for Riverland Government ``` Now develop influence tactics for Riverland Government using the latest analysis. ``` **Response:** Specific tactics for Riverland Government, including coalition opportunities (Donors + Local NGOs identified as allies). ### Step 2b: Develop Tactics for GRN (Armed Group) ``` Now develop influence tactics for GRN using the latest analysis. ``` **Response:** Different tactics for GRN, recognizing them as adversarial (position: -0.8). ### Step 2c: Develop Tactics for Other Actors Repeat for Donors, Media, Local NGOs - just change the name! --- ## Complete Tool Set Available | # | Tool | Purpose | Input | |---|------|---------|-------| | **1** | Island of Agreement | Understand common ground | Situation + two parties | | **2** | Icebergs | Dig into interests | Positions, reasoning, motives | | **3** | Stakeholders | Prioritize actors | List of stakeholders with ratings | | **4** | Leverage Influence (OLD) | Tactics (manual) | Name + full analysis data | | **5** | **Leverage Influence Latest** | **Tactics (automatic!)** | **Just the stakeholder name** ⭐ | | **6** | Guide | Reference | No input needed | --- ## Why This Works Better **Before (Old Way):** ``` Tool 3: Analyze Stakeholders ↓ Get response ↓ Copy entire JSON ↓ Tool 4: Leverage Influence ↓ Paste JSON + target name ↓ Get tactics ``` **Now (New Way):** ``` Tool 3: Analyze Stakeholders ↓ Get response ↓ Tool 5: Leverage Influence Latest ↓ Just say target name ↓ Get tactics (automatically uses analysis) ``` --- ## Setting Up in ChatGPT 1. Go to your Custom GPT → **Actions** 2. Replace your schema with: `OPENAI_COMPLETE_SCHEMA.json` 3. Server URL: `https://humanitarian-mcp-904769523449.us-central1.run.app` 4. You'll now have 6 tools available (including the new auto one!) --- ## Multiple Analyses in Session The system keeps the **latest stakeholder analysis for 30 minutes**. If you run a new stakeholder analysis, it replaces the old one. - Run analysis for Riverland crisis - Develop tactics for 5 stakeholders - Then run analysis for different crisis - Develop tactics for new stakeholders - The system automatically switches to the new analysis --- ## Error Messages & Solutions | Error | Cause | Solution | |-------|-------|----------| | "Stakeholder 'X' not found" | Name doesn't match exactly | Use exact name from analysis | | "No cached analysis available" | You haven't run Tool 3 yet | Run stakeholder analysis first | | "Cached analysis expired" | More than 30 min passed | Run stakeholder analysis again | --- ## Recommended Workflow for Complex Negotiations ``` 1. Island of Agreement → Understand where you agree 2. Icebergs (optional) → Understand motivations 3. Stakeholders → Prioritize who matters most 4. Leverage Influence Latest (repeat for each key stakeholder) → Get specific tactics 5. Guide (reference) → Review best practices ``` --- ## Pro Tips - **Develop tactics for ALL stakeholders**, not just friendly ones - GRN (adversarial) tactics are just as important - Media (neutral) tactics help shape narrative - **Update analysis after negotiation events** - Stakeholder positions change - Re-run analysis with new data - Develop new tactics accordingly - **Use before/after analysis** - Run analysis before negotiation → develop tactics - Run analysis after → see what changed - Adjust strategy based on position changes --- ## Available Endpoints All endpoints on: `https://humanitarian-mcp-904769523449.us-central1.run.app` ``` POST /api/v1/island-of-agreement POST /api/v1/analyze-icebergs POST /api/v1/analyze-stakeholders ← Feeds into Tools 4 & 5 POST /api/v1/leverage-influence ← Manual: requires full analysis data POST /api/v1/leverage-influence-latest ← Automatic: uses cached data ⭐ NEW! GET /api/v1/guide ``` --- ## Support Deployment status: ✅ Always online on Google Cloud Run Health check: `https://humanitarian-mcp-904769523449.us-central1.run.app/health` All tools tested and working! 🎉

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/escenariosparalatransformacion/humanitarian-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server