content-seo

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Instructions in SKILL.md direct the agent to execute shell commands (ls) to probe the local filesystem for internal installation directories (~/.claude/skills/, ~/.agent/skills/, etc.) to detect environment state.
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to offer the installation of additional remote code via npx skills add from an unverified external source (AbsolutelySkilled/AbsolutelySkilled).
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it processes untrusted SEO content without security boundaries. 1. Ingestion points: SEO content briefs and user drafts provided during audits (SKILL.md, references/content-gap-analysis.md). 2. Boundary markers: Absent; no instructions are provided to distinguish between data and instructions. 3. Capability inventory: Shell execution capabilities (ls) and the ability to suggest external installations. 4. Sanitization: Absent; the skill does not define validation or filtering for ingested content.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 03:03 PM
Security Audit — agent-trust-hub — content-seo