lov-better-seo
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data such as website metadata,
sitemap.xml, androbots.txtfiles to perform SEO audits. - Ingestion points: Target project files and public web resources (Workflow steps 1 and 4 in
SKILL.md). - Boundary markers: No specific technical delimiters are used, but the skill defines a policy-based execution boundary that keeps analysis and preview actions strictly read-only.
- Capability inventory: The skill possesses file-system read/write capabilities for project configuration and executes a local Python script (
profile_store.py) to manage its state. - Sanitization: No explicit sanitization of ingested content is described; however, the skill focuses on structural and semantic SEO checks rather than executable logic.
- [DATA_EXPOSURE]: The skill utilizes a persistence mechanism via
scripts/profile_store.pyto store user-stated preferences and brand facts in a local JSON profile. - Evidence: The script includes a blocklist (
SENSITIVE_PARTS) that prevents writing data to paths containing sensitive keywords like 'token', 'secret', or 'api_key'. - Context: This storage is restricted to
~/.config/agent-skills/or similar user-owned directories and uses atomic write operations to prevent file corruption.
Audit Metadata