The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill makes network requests to well-known security services such as Have I Been Pwned and BreachDirectory to check for leaked data. It also fetches ransomware incident information from the Ransomwatch repository on GitHub. These sources are considered well-known technology services and the references are documented neutrally as they align with the skill's stated purpose.
[PROMPT_INJECTION]: The skill is designed to ingest and process text from external, untrusted sources including dark web paste sites and ransomware leak forums. This creates a surface for indirect prompt injection, where malicious instructions embedded in the collected threat intelligence could potentially influence the behavior of an AI agent consuming the output.
Ingestion points: Content is retrieved from onion paste sites in 'SKILL.md' and ransomware leak data in 'scripts/process.py'.
Boundary markers: The scripts do not implement delimiters or 'ignore instructions' warnings for the ingested data.
Capability inventory: The agent scripts perform network operations via the requests library and write findings to local JSON files in 'scripts/agent.py' and 'scripts/process.py'.
Sanitization: There is no evidence of escaping or validation of the external content before it is interpolated into reports or displayed to the user.

performing-dark-web-monitoring-for-threats