tests-prod-test-bitly-links
Fail
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: CRITICALCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill employs
python3 -cto decode URL anchors using a command string that interpolates the${FINAL_URL#*#}shell variable. BecauseFINAL_URLis the result of following an external redirect, a maliciously crafted URL anchor could break out of the Python string and execute arbitrary code within the Python sub-process. - [EXTERNAL_DOWNLOADS]: The skill is configured to perform network requests to
itautonomos.comand follow variousbit.lyredirects. Automated security scans have identifiedhttps://bit.ly/example, which is used as a placeholder in the skill's documentation, as a malicious URL associated with botnet activity. - [DATA_EXPOSURE]: The skill accesses the local file
_resources/bitly_links.jsonto validate link coverage. This involves reading internal project resources which, while necessary for the task, provides the agent with access to local filesystem data.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata