Instructions file
Instructions for hoophq/fence, covering claude.md, project overview, commands, architecture — agent-neutral core + per-agent adapters and conventions (load-bearing).
19 29d ago E 1,815 tokens
original MIT
Guardrails for AI coding agents — block catastrophic tool calls before they run. Semantic, agent-neutral, near-zero false positives.
Instructions file
Instructions for hoophq/fence, covering claude.md, project overview, commands, architecture — agent-neutral core + per-agent adapters and conventions (load-bearing).