board
Warn
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill documentation suggests a pattern of executing shell commands where user-controlled variables are interpolated into a string:
python {skill_path}/scripts/board_manager.py --post --channel {channel} --author {author} --message "{text}". This approach is vulnerable to shell command injection if the variables contain metacharacters like semicolons, backticks, or command substitutions.\n- [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface (Category 8) because it reads and displays content from a shared message board that can be influenced by other agents or external inputs.\n - Ingestion points: The agent reads posts via the
board_manager.py --readcommand, bringing external data into its active context.\n - Boundary markers: There are no instructions or delimiters provided to help the agent distinguish between its own system instructions and the content of the messages it reads.\n
- Capability inventory: The skill allows the agent to write files and execute local scripts, which could be abused if the agent is manipulated by a malicious post.\n
- Sanitization: The skill does not describe any process for sanitizing the input from the message board before the agent processes it.
Audit Metadata