MCP server developer: same tools behave differently across Claude, ChatGPT, Gemini
quiet_flare · reddit · 2026-11-19
A developer running a public MCP server shared observations: the same tools behave differently when called via Claude, ChatGPT, Gemini, and Grok, with failures mostly from models picking the wrong tool or receiving more data than they can use.
He asked peers four practical questions:
- Do you measure wrong-tool/wrong-argument rates? How, and what numbers do you see?
- Did you start from an API-spec-generated server with one tool per endpoint, and what changed after real users arrived?
- Has the same server worked in one assistant but failed in another — and where exactly?
- Was building a connector customer-driven or your own roadmap call?
More from coding & agent
- Dev builds openclaw-style agent tool usable from any or multiple harnesses — cgarciae88 · 2026-11-19
- Dev marks one year since writing his last line of code by hand, credits AI — Dimillian · 2026-11-19
- FlowX: open-source compiler turns agent experience into self-evolving workflows — Over_Accountant_2311 · 2026-11-19
- Dev builds strict verification harness to stop coding agents from faking green tests — Gold_Exchange6103 · 2026-11-19
- Univer hits #1 on GitHub Trending with 140+ AI workflow scenarios across 25+ industries — alifcoder · 2026-11-19
- nohumans.directory bought 8,000+ paid x402 APIs with real USDC so your agent knows before it pays — SashSail · 2026-11-19