content-atomizer

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of processing untrusted external data.
  • Ingestion points: The skill reads long-form content from user input, blog posts, newsletters, and transcripts located in the marketing/content/ directory as described in SKILL.md.
  • Boundary markers: The instructions do not define clear delimiters or "ignore instructions" warnings to isolate the source content from the agent's logic, which could allow embedded commands in the source text to influence agent behavior.
  • Capability inventory: The skill has the capability to write files to the local file system (using the source-slug variable) and suggests the use of playwright-cli for automated social media posting in references/platform-specs.md.
  • Sanitization: No sanitization or validation of the input content (including the title used for the source-slug) is specified, which could lead to unintended output or path manipulation if not handled by the underlying agent platform.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 02:06 PM
Security Audit — agent-trust-hub — content-atomizer