attack-path-architect
Warn
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill's instructions require the agent to execute Python scripts using shell commands with direct string interpolation of user-controlled variables (e.g.,
python scripts/classify_assets.py --input {recon_data}). This creates a significant risk of command injection if variables such as{recon_data},{objective}, or{name}contain shell metacharacters. - [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection because it ingests and processes external reconnaissance data which could be manipulated by an attacker to include hidden instructions.
- Ingestion points: Data provided as JSON or domain info in Step 1 and processed in Step 2 (
recon_data). - Boundary markers: None. There are no delimiters or instructions to treat the data as inert content.
- Capability inventory: The agent can execute shell scripts, generate reports, and map attack paths based on this data.
- Sanitization: No validation or sanitization of the external data is performed before it is used in logic or command execution.
Audit Metadata