twitter-thread
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified in the skill. The instructions are focused on creative writing and formatting for social media.
- [DATA_EXPOSURE]: The skill references local context files located in
/context/(e.g.,voice-dna.json,icp.json,business-profile.json) and content from/knowledge/. This is standard behavior for skills designed to maintain brand voice and context and does not involve unauthorized data access or exfiltration. - [PROMPT_INJECTION]: The skill does not contain instructions that attempt to override AI safety filters or hijack the agent's core personality. It uses natural instructional language to guide the content generation process.
- [REMOTE_CODE_EXECUTION]: There are no patterns involving package managers, shell command execution, or remote script downloads.
- [INDIRECT_PROMPT_INJECTION]: While the skill ingests external data (user content to be repurposed), it lacks the capabilities (such as shell access or network requests) that would make such an injection dangerous.
- Ingestion points: Reads content from
/knowledge/and user-provided long-form content (documented in SKILL.md). - Boundary markers: None explicitly defined in the prompt template.
- Capability inventory: Limited to text generation and formatting.
- Sanitization: No specific sanitization or validation of the input content is mentioned.
Audit Metadata