archive-conversation
Pass
Audited by Gen Agent Trust Hub on Mar 23, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted conversation history to generate summaries.
- Ingestion points: The 'Processing Logic' in SKILL.md requires the agent to analyze the full conversation history.
- Boundary markers: No delimiters or protective instructions are provided to separate conversation data from agent instructions.
- Capability inventory: The skill utilizes 'Write', 'Glob', and 'Bash' tools.
- Sanitization: There are no instructions to sanitize or escape data before writing it to the file system.
- [COMMAND_EXECUTION]: The skill allows the user to specify an output-path for the archive. Without explicit path validation or restriction instructions, an attacker could potentially induce the agent to perform path traversal or overwrite sensitive files using the 'Write' tool.
Audit Metadata