Agent Cards v0.1 draft. Declarative JSON document disclosing an AI agent's capability surface, refusal behavior, evaluations, and deployment posture. Like HuggingFace model cards, but for agents. Part of the Kinetic Gain Protocol Suite.

Open 30-attack prompt-injection corpus + Python harness with Agent Card refusal-taxonomy back-references. 10 attack categories, 4-level severity, 5 scoring methods. Pairs with the Kinetic Gain Protocol Suite.

Visual web harness for prompt-injection-bench. Paste a JSONL transcript, see a procurement-grade pass-rate report. Live at bench.kineticgain.com. Client-side React + TS.

Unified MCP server exposing all 5 Kinetic Gain Protocol Suite specs as tools (18 total across AEO, Prompt Provenance, Agent Cards, AI Evidence, MCP Tool Cards). One Claude Desktop config entry. Companion to kinetic-gain-visualizer.