Agentic Loop Memory Server
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the "@mhrj/mcp-agent-loop" package from the author's namespace to provide the core memory server functionality.
- [COMMAND_EXECUTION]: The setup instructions involve running the "npx" command to execute the remote MCP server package.
- [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection through its state management system. Ingestion points: Untrusted data from command outputs and file edits is recorded into the state file via the "log_step" tool in "SKILL.md". Boundary markers: The skill does not define specific delimiters or instructions in the state file to prevent the agent from obeying instructions embedded in the logged results when the state is read back. Capability inventory: The agent is authorized to execute bash commands and modify files across all scripts. Sanitization: No evidence of sanitization or filtering of the content logged to the memory state was found.
Audit Metadata