tree-ring-memory-cn
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions include several shell command examples using a CLI tool named
tree-ring. This includes subcommands for recalling memory, syncing documentation (dox sync), and scanning integrations. - [PROMPT_INJECTION]: The skill explicitly directs the agent to read and follow instructions from local project files:
.tree-ring/SKILL.mdand.tree-ring/CLI.md. This creates a surface for indirect prompt injection where a malicious repository could include these files to override the agent's behavior or trigger unauthorized tool use when the repository is opened. - Ingestion points: Local files located at
.tree-ring/SKILL.mdand.tree-ring/CLI.mdrelative to the current working directory. - Boundary markers: Absent. The skill does not provide instructions to delimit or treat the content of these local files as untrusted data.
- Capability inventory: The agent is provided with command execution capabilities via the
tree-ringCLI and potentially other tools configured in the agent's environment. - Sanitization: Absent. There are no instructions for the agent to validate or sanitize the instructions found in the local configuration files before adopting them.
Audit Metadata