competitor-monitoring

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill fetches and processes content from the live web, including competitor blogs, news articles, and changelogs, creating an ingestion point for untrusted instructions.
  • Ingestion points: Web search queries and full-page text extraction for priority signals (defined in SKILL.md under 'Fan out the research' and references/query-patterns.md under 'Changelog extraction').
  • Boundary markers: Absent. The instructions do not specify the use of delimiters or warnings to separate external content from the agent's core instruction set.
  • Capability inventory: The skill utilizes web search and extraction tools, and maintains read/write access to a local workspace directory (competitive-intel/) to store profiles, briefings, and memory files.
  • Sanitization: The validation process focuses on data integrity (event dates and source corroboration) as described in references/signal-validation.md, but lacks specific sanitization or filtering for embedded instructions in the ingested content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 06:57 AM
Security Audit — agent-trust-hub — competitor-monitoring