openclaw-migration
Pass
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: The migration script accesses local configuration files and environment variables containing sensitive data (e.g., API keys) within the ~/.openclaw directory. This access is limited to the local migration process. The script includes functionality to redact these secrets from the generated migration reports to prevent accidental exposure in logs.
- [PROMPT_INJECTION]: The skill ingests data from local files into the agent's context, which represents a surface for indirect prompt injection.
- Ingestion points: Local MEMORY.md, USER.md, and SOUL.md files.
- Boundary markers: Memory entries are delimited by the "§" character.
- Capability inventory: The script can modify agent configuration and local environment variables.
- Sanitization: It performs branding-related text replacement but lacks sanitization for embedded natural language instructions.
- [COMMAND_EXECUTION]: The skill invokes a Python script to perform the migration. Security is maintained through a mandatory dry-run requirement and the use of the clarify tool for interactive user approval of all destructive or configuration-altering actions.
Audit Metadata