The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documentation includes numerous examples for executing the nuclei CLI tool and associated pipeline tools like subfinder and httpx to perform network scanning and vulnerability assessment.
[EXTERNAL_DOWNLOADS]: Includes standard commands for updating vulnerability templates from official sources using the nuclei -update-templates flag. These templates are maintained by a well-known security community.
[REMOTE_CODE_EXECUTION]: Documents the capability to execute vulnerability templates directly from remote URLs via the -t <url> flag, which is a core feature of the documented tool for testing specific exposures.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion and processing of data from external network targets during scans, which could theoretically contain malicious payloads intended for the analyzing agent.
Ingestion points: Network responses from scanned targets, input files containing URLs/hosts, and scan results in JSONL format (e.g., findings.jsonl).
Boundary markers: Not explicitly used; the skill relies on standard shell piping and redirection.
Capability inventory: nuclei, subfinder, httpx, cat, jq, sort, uniq used for processing scan data.
Sanitization: No specific sanitization or filtering is applied to the untrusted output of the scan tools before it is processed by utilities like jq.

nuclei