The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external websites via tools like inspect_site_tracking and inspect_runtime_consent to analyze tracking signals. This creates an attack surface where a malicious website could include hidden instructions aimed at biasing the analysis or manipulating the final report. * Ingestion points: External data from user-supplied URLs is processed and integrated into the agent context within SKILL.md. * Boundary markers: The instructions do not define specific delimiters or warnings to ignore embedded instructions within the ingested site data. * Capability inventory: The skill uses export_consent_report to generate file content based on the external observations, allowing potential injection content to persist in the generated output. * Sanitization: No explicit sanitization or filtering of the retrieved site content is mentioned to mitigate malicious prompts.
[EXTERNAL_DOWNLOADS]: The skill directs the agent to utilize external tools and documentation hosted at https://regsentry.com/mcp and https://regsentry.com/mcp-guide for its core workflow.

regsentry-consent-evidence-review