edge-profiler
Pass
Audited by Gen Agent Trust Hub on Mar 30, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill instructions are focused on structured data extraction and do not contain any patterns intended to bypass safety filters, override agent personality, or leak system prompts.
- [DATA_EXFILTRATION]: No network tools or external URLs are present in the skill files. The workflow operates entirely within the LLM's context window without transferring data to external servers.
- [COMMAND_EXECUTION]: No shell commands, CLI tools, or subprocess executions are defined or requested. The skill is purely informational.
- [REMOTE_CODE_EXECUTION]: The skill does not use package managers or download scripts from the internet. All logic is contained within the markdown and JSON files provided.
- [CREDENTIALS_UNSAFE]: No hardcoded credentials or sensitive file paths were found. The skill specifically instructs the agent to move employer-confidential information into a 'restricted' category, demonstrating a security-conscious design.
- [EXTERNAL_DOWNLOADS]: All references within the skill are to local files (e.g., assets/profile-template.json, references/profile-dimensions.md). No external resources are fetched.
Audit Metadata