Building Blocks
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches content from
https://ethskills.com/building-blocks/SKILL.mdusingcurlas part of its installation and setup process. - [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read external content from
https://ethskills.com/building-blocks/SKILL.mdbefore performing DeFi operations. This creates a surface for indirect prompt injection where malicious instructions could be embedded in the remote file to influence agent behavior. - Ingestion points: External URL
https://ethskills.com/building-blocks/SKILL.mdaccessed during setup and operation (specified incatalog.json). - Boundary markers: None present in the instructions to distinguish external data from core instructions.
- Capability inventory: The agent's primary purpose is to "build DeFi," which involves high-privilege activities such as generating smart contract code or transaction parameters.
- Sanitization: No evidence of sanitization, filtering, or validation of the fetched remote content before it enters the agent context.
- [REMOTE_CODE_EXECUTION]: The setup instructions in
catalog.jsonrecommend installing a plugin from a third-party GitHub repository (https://github.com/austintgriffith/ethskills) and using theclawhubpackage manager, which involves the installation and execution of external code from sources outside of the trusted list.
Audit Metadata