name: camel-guard repo: https://github.com/nativ3ai/hermes-camel-guard-plugin sha: 4cbfca3a5f1d46f5d152280cccc496d0df14d097 subdir: plugin title: CaMeL Guard description: Gate sensitive tools on trusted user capabilities after untrusted tool output. maintainer: nativ3ai tier: community category: tools requires_hermes: ">=0.22.7" version: "1.1.1" docs_url: https://github.com/nativ3ai/hermes-camel-guard-plugin/blob/4cbfca3a5f1d46f5d152280cccc496d0df14d097/README.md readme: true known_issues: - Coarse capability authorization does not constrain individual tool arguments. - Enforce blocks sensitive actions when the configured model cannot classify intent. capabilities: provides_tools: [] provides_hooks: - pre_llm_call - pre_tool_call - post_tool_call - on_session_end - on_session_reset provides_middleware: [] requires_env: []