linkedin-post-generator
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: No security issues were detected. The skill performs its stated function of generating social media content based on user input and local configuration files.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided material (notes, stories, topics) and reads a configuration file (~/.liftli/voice.md). This is an inherent part of its design as a writing tool and does not present a high risk as the skill does not have access to sensitive system tools or perform automated high-privilege actions. 1. Ingestion points: User chat inputs and .liftli/voice.md / ~/.liftli/voice.md files (SKILL.md). 2. Boundary markers: None defined. 3. Capability inventory: Call to strategist_briefing tool if available. 4. Sanitization: None defined.
- [EXTERNAL_DOWNLOADS]: The skill documentation suggests user-initiated installation of related tools using npx and claude mcp commands targeting the developer's official repository (liftli-ai/linkedin-agent-skills) and domain (liftli.ai). These are standard platform extension mechanisms and are correctly aligned with the vendor's identity.
Audit Metadata