The Agent Skills Directory

[SAFE]: The skill functions as a legitimate data extraction tool for gathering speaker information from public conference websites and does not exhibit any malicious patterns.
[PROMPT_INJECTION]: The skill possesses a potential surface for indirect prompt injection, as it processes content from untrusted external URLs that could contain adversarial text aimed at the agent.
Ingestion points: The scripts/scrape_speakers.py script fetches data from user-provided URLs using requests.get or the Apify proxy.
Boundary markers: No specific delimiters or markers are applied to the scraped data before it is returned to the agent.
Capability inventory: The skill is capable of performing network operations via the requests library.
Sanitization: The script includes logic to strip HTML tags via regular expressions but does not perform sanitization to detect or neutralize prompt injection attacks within the extracted text.
[DATA_EXFILTRATION]: When the Apify mode is active, the skill sends the target URL to the vendor's API endpoint at app.gooseworks.ai. This is expected behavior for the skill's optional proxy scraping functionality.
[CREDENTIALS_UNSAFE]: The skill manages its API key through the GOOSEWORKS_API_KEY environment variable. This follows standard security practices for managing secrets in scripts.

conference-speaker-scraper