ghchinoy/agent-seer

Specification-driven evaluation, deterministic capability linting, and synthetic test harness generation for Model Context Protocol (MCP) tool suites. Validates tool-calling transcripts against schemas and runtime capability matrices in <1ms without LLMs, and scores semantic orchestration using a decomposed 14-subdimension LLM-as-judge rubric. Use when testing agent tool-calling reliability, catching model-capability footguns, linting MCP function calls, or generating synthetic multi-turn eval harnesses from tools/list specs.