social-calendar-builder
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill manages a potential attack surface by ingesting external data, but incorporates specific defensive instructions to mitigate risks.
- Ingestion points: The skill processes potentially untrusted data from analytics exports, trend reports, and scraped pages as described in the instructions section of
SKILL.md. - Boundary markers: Explicit security instructions are included in
SKILL.mdthat command the agent to treat external data as untrusted and prevent it from overriding committed settings or status markers. - Capability inventory: The skill interacts with the file system (
memory/paths) and uses a dedicated script (registry-events.py) to propose updates to the channel registry. - Sanitization: The skill relies on natural language instructions to sanitize the impact of embedded commands within the processed data.
Audit Metadata