The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/parallel_sector_probe.py utilizes the Python subprocess module to execute system commands. It specifically wraps the curl utility to perform automated network requests and vulnerability probing against target domains.
[EXTERNAL_DOWNLOADS]: The skill instructions and associated Python script perform network requests to crt.sh, an established Certificate Transparency search engine, to retrieve domain names for specific industry sectors for the purpose of target list compilation.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing external data retrieved from certificate transparency logs and web server responses.
Ingestion points: Untrusted data enters the agent context via the crt.sh API results and raw HTTP responses (including JSON user data and headers) processed by both the shell commands in SKILL.md and the scripts/parallel_sector_probe.py script.
Boundary markers: There are no explicit boundary markers or instructions to ignore embedded commands when the agent processes these external target lists or probe results.
Capability inventory: The skill has the capability to execute shell commands via subprocess.run and perform file writing operations to store findings.
Sanitization: The skill uses standard utilities like jq, sed, and grep to filter domain strings from JSON logs, and uses json.loads in Python for parsing user enumeration results.

sector-recon-methodology