The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script scripts/event_signals.py to automate the collection and processing of event data from various platforms.
[EXTERNAL_DOWNLOADS]: The skill fetches data from several external services including Sessionize, Confs.tech, Apify (proxied through the vendor domain api.gooseworks.ai), ListenNotes, and Devpost to gather speaker and event information.
[PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection.
Ingestion points: Untrusted content is retrieved from conference websites and external APIs as defined in SKILL.md and scripts/event_signals.py.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the ingested external content as untrusted or to ignore embedded instructions.
Capability inventory: The skill environment allows access to powerful tools like Bash, WebFetch, and WebSearch which could be exploited if malicious instructions are present in the external data.
Sanitization: The scripts/event_signals.py file contains a clean_html function to sanitize data for CSV storage, but no semantic sanitization is applied to content before it is processed by the agent.

event-signals