fixing-broken-links
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
curlcommands to test the validity of external URLs. While necessary for the task, executing shell commands with strings extracted from untrusted files poses a risk of command injection if the agent does not properly escape the input before passing it to the shell. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from user-provided files which can influence agent behavior. An attacker could craft specific patterns in the target files to mislead the agent or exploit its capabilities.
- Ingestion points: Target files and projects provided for scanning (SKILL.md).
- Boundary markers: None specified in the instructions to separate extracted data from instructions.
- Capability inventory: Shell execution of
curl, file system checks ([ -f ]), and file writing operations for link replacement. - Sanitization: No explicit sanitization or validation of extracted URLs is mentioned before they are processed.
- [DATA_EXFILTRATION]: The skill reads local files and performs network requests. This creates a potential path for sensitive data to be exfiltrated if a malicious file contains URLs that embed sensitive information as query parameters, which the skill then 'tests' by sending a request to an external server.
Audit Metadata