The Agent Skills Directory

[SAFE]: No security concerns were identified. The file contains only descriptive metadata and high-level workflow instructions.
[NO_CODE]: There are no executable scripts, commands, or code files included in this skill.
[PROMPT_INJECTION]: The skill represents a potential surface for indirect prompt injection as it is designed to monitor and ingest external research data. Ingestion points: External research topics, papers, and product updates. Boundary markers: None specified in the metadata. Capability inventory: Utilizes a platform 'researcher' agent. Sanitization: No explicit sanitization or validation steps for external content are defined in this configuration file.

watch