Use AI to explore Bittensor
Point any agent at metagraphed — over MCP, a typed SDK, or plain HTTP — and it can find, explain, and call the right Bittensor subnet for a task. No key, no account.
Connect over MCP
One command in Claude Code, Cursor, or any MCP client. 17 tools over streamable-http — search the registry, find a subnet for a task, get a callable RPC endpoint, ask a grounded question.
claude mcp add --transport http metagraphed https://api.metagraph.sh/mcpsearch_subnets · list_subnets · find_subnets_by_capability · get_subnet · get_subnet_health · list_subnet_apis · get_api_schema · get_fixture · get_agent_catalog · get_best_rpc_endpoint · registry_summary · list_enrichment_targets · semantic_search · ask · find_subnet_for_task · how_do_i_call · verify_integration
Or install the SDK
Typed clients for Python and TypeScript that wrap every route and the RPC proxy.
pip install metagraphedPython · metagraphednpm i @jsonbored/metagraphedTypeScript · @jsonbored/metagraphedOr drop into a chat
Paste-ready system prompt that turns any agent (Claude, Cursor, …) into a metagraphed-powered Bittensor integration agent.
system prompt api.metagraph.sh/agent.md
Everything else, fetchable directly
A paste-ready agent prompt, a Bittensor skill, llms.txt, the OpenAPI contract, grounded Q&A, semantic search, and bulk data — 92 callable services across 129 subnets, all indexed at /api/v1/agent-resources.
Try it
No key, no account — hit any surface with curl.
curl -s https://api.metagraph.sh/api/v1/ask \
-X POST -H 'content-type: application/json' \
-d '{"question":"which subnet does image generation?"}'curl -s https://api.metagraph.sh/api/v1/agent-catalog
curl -s 'https://api.metagraph.sh/api/v1/search/semantic?q=video+generation'