Trust-gradient skills for Claude Code. Capability compounds: skills go .proposed → .verified → .trusted → .autonomous through demonstrated reliability. No daemon, no cloud.
python skills memory ai-agents local-first self-improving anthropic agent-tools claude-code claude-code-plugin claude-plugin verified-ai trust-gradient compounding-trust no-daemon
-
Updated
May 11, 2026 - Python