Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Discover and search 18K+ MCP servers and AI agents across 6+ registries using Global Chat's cross-protocol directory and MCP server.
.claude/skills/global-chat-agent-discovery/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-19 | ✗→✓ | ▲ Improved | — | — |
| case-13 | ✗→✓ | ▲ Improved | — | — |
| case-08 | ✗→✓ | ▲ Improved | — | — |
| case-20 | ✗→✓ | ▲ Improved | — | — |
| case-11 | ✗→✓ | ▲ Improved | — | — |
Global Chat is a cross-protocol AI agent discovery platform that aggregates MCP servers and AI agents from 6+ registries into a single searchable directory. This skill helps you find the right MCP server, A2A agent, or agents.txt endpoint for any task by searching across 18,000+ indexed entries. It also provides an MCP server (@global-chat/mcp-server) for programmatic access to the directory from any MCP-compatible client.
Install the Global Chat MCP server to search the directory programmatically from Claude Code, Cursor, or any MCP client.
bashnpm install -g @global-chat/mcp-server
Add to your MCP client configuration:
json{ "mcpServers": { "global-chat": { "command": "npx", "args": ["-y", "@global-chat/mcp-server"] } } }
Then ask your agent to search for tools:
Search Global Chat for MCP servers that handle PostgreSQL database queries.Browse the full directory at https://global-chat.io:
If you maintain an agents.txt file, use the free validator:
You: "Find MCP servers that can convert PDF files to text"
Agent (via Global Chat MCP): Searching across 6 registries...
- @anthropic/pdf-tools (mcpservers.org) — PDF parsing and text extraction
- pdf-converter-mcp (mcp.so) — Convert PDF to text, markdown, or HTML
- ...You: "What A2A agents are available for code review?"
Agent (via Global Chat MCP): Found 12 A2A agents for code review across 3 registries...You: "How many registries list tools for Kubernetes management?"
Agent (via Global Chat MCP): 4 registries carry Kubernetes-related agents (23 total entries)...Solution: Add protocol or registry filters to narrow the scope
Solution: Ensure npx is available and run npx -y @global-chat/mcp-server manually first to verify
@mcp-client - For general MCP client setup and configuration@agent-orchestration-multi-agent-optimize - For orchestrating multiple discovered agents@agent-memory-mcp - For persisting discovered agent information across sessions| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-19 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +70 percentage points is the difference between those two pass rates over the 23 comparable cases.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.