update-nanoclaw
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill downloads update logic from the official vendor repository at
https://github.com/nanocoai/nanoclaw.gitand executes it. This is the primary intended function for managing self-updates of the installation.\n- [DYNAMIC_EXECUTION]: Scripts are fetched into a temporary directory and executed usingpnpm exec tsx. The skill uses a staged approach to validate these scripts before applying changes to the live environment.\n- [INDIRECT_PROMPT_INJECTION]: The skill is instructed in Step 5 to read and followSKILL.mdfiles from other installed skills during the migration phase.\n - Ingestion points: Step 5 reads external
SKILL.mdfiles to process migration requirements.\n - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' warnings for the ingested content.\n
- Capability inventory: The skill has significant capabilities including shell execution via
git,pnpm, andtsx, as well as service management and file system snapshots.\n - Sanitization: There is no mention of sanitizing or escaping content from the external skill files before processing.
Audit Metadata