nuclei-scan
Fail
Audited by Gen Agent Trust Hub on May 9, 2026
Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The shell command execution in Step 3 is vulnerable to injection. Variables such as $TARGET_OPT, $SEVERITY, $TMPL_OPT, and $RATE are interpolated into the nuclei command without shell quoting. A malicious user providing input via environment variables like SECSKILL_TARGET can include shell metacharacters (e.g., ; or &&) to execute arbitrary commands on the host.
- [EXTERNAL_DOWNLOADS]: The skill installs the Nuclei binary and updates vulnerability templates from ProjectDiscovery's official GitHub repository.
- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by processing results from external network targets.
- Ingestion points: Vulnerability scan results are read from the JSON output file in Steps 5 and 6 of SKILL.md.
- Boundary markers: Absent. There are no delimiters or instructions to ignore content within the ingested data.
- Capability inventory: The skill has capabilities for shell execution (via nuclei) and filesystem writes (for result logs).
- Sanitization: Content from the scan results is parsed as JSON and printed to the console by an inline Python script without validation or escaping.
- [DATA_EXFILTRATION]: Performs outbound network requests to scan external targets and fetch tool updates.
Recommendations
- AI detected serious security threats
Audit Metadata