Verifiable deep research for anything onchain. Reasoning by SERV.
SERV: connected (gpt-5.4-mini) | Skill: loaded (28 files)
MCP endpoint (streamable HTTP):
https://pima5-seeri.hf.space/mcp
claude mcp add --transport http seeri https://pima5-seeri.hf.space/mcp
deep_research | Full pipeline: decompose, gather, classify, grade, synthesize (markdown) |
deep_research_json | Same pipeline, structured JSON output |
verify_claim | Grade a single claim against gathered evidence |
robinhood_token_dossier | Contract identity, ticker collisions, pools on Robinhood Chain |
protocol_snapshot | DefiLlama TVL, chains, category for a protocol or chain |
web_evidence | Web search with classified source list |
seeri_skill | Read a Seeri skill file (the method) |
list_skill_files | List every file in the loaded skill |
Educational research output. Not investment, legal, or tax advice.