pdb-structure
Warn
Audited by Gen Agent Trust Hub on Aug 18, 2026
Risk Level: MEDIUMCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill instructions advise setting the FITTINGS_PRIVATE_KEY environment variable with a blockchain private key. Storing sensitive credentials in plaintext environment variables is a high-risk practice as they may be exposed to the AI agent or leaked through process metadata.
- [EXTERNAL_DOWNLOADS]: The documentation includes a curl command to download the execution script call.mjs from the vendor's domain (fittings.sh). This introduces a runtime dependency on externally hosted code.
- [REMOTE_CODE_EXECUTION]: The execution workflow involves downloading a remote script and running it via node. This pattern allows for the execution of unverified code that could be modified on the remote server.
- [COMMAND_EXECUTION]: The skill utilizes shell commands (node, curl) to perform its functions, including making network requests and signing blockchain transactions.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from a remote API which could influence agent behavior. 1. Ingestion points: API responses from fittings.sh processed in call.mjs. 2. Boundary markers: None identified in instructions or scripts. 3. Capability inventory: Network requests via fetch, cryptographic signing, and command execution via node. 4. Sanitization: No data validation or prompt-level sanitization is implemented for the external API content.
Audit Metadata