The Agent Skills Directory

[SAFE]: The core logic is implemented in scripts/probe_site.py, which uses standard Python libraries to perform controlled network probes. The script identifies security headers, cookies, and challenge scripts to characterize website defenses without attempting to bypass them.
[SAFE]: The skill instructions in SKILL.md explicitly define boundaries, excluding authenticated content and the evasion of explicit blocks. It prioritizes sanctioned data access methods, such as official APIs and data exports, which reduces the need for adversarial scraping.
[SAFE]: The documentation provided in the references/ directory contains technical signatures for legitimate security vendors and lists reputable open-source libraries (e.g., httpx, Playwright, Scrapy).
[SAFE]: The skill ingests external data from robots.txt and page HTML via a Python sub-process. While this presents a surface for indirect prompt injection, it is a necessary part of the auditing function. The instructions mitigate this by directing the agent to analyze the data as structured evidence rather than following instructions contained within the fetched content.

scrape-feasibility-audit