fittings-regulations-document-search
Warn
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: MEDIUMCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill instructions direct the user to store a sensitive cryptocurrency private key in the
FITTINGS_PRIVATE_KEYenvironment variable. Storing private keys in plaintext environment variables poses a risk of credential exposure to other processes, shell history, or logs. - [EXTERNAL_DOWNLOADS]: The skill retrieves an external JavaScript file (
call.mjs) from the developer's domain (https://fittings.sh/skill/call.mjs). - [REMOTE_CODE_EXECUTION]: The skill executes the downloaded script using the command
node call.mjs. This involves running external code at runtime to interact with the API, which is a high-risk pattern if the source is compromised. - [COMMAND_EXECUTION]: The documentation provides shell commands that perform network downloads (
curl), set environment variables (export), and execute scripts (node). - [INDIRECT_PROMPT_INJECTION]: The skill ingests rulemaking data from an external government API via the fittings.sh gateway, creating a potential surface for indirect prompt injection if the data contains malicious instructions. No specific sanitization or boundary markers are documented for the processed content.
- Ingestion points: Fetches search results from Regulations.gov via the
GET /v1/regulations/document-searchendpoint. - Boundary markers: None mentioned in the instructions for handling the external data.
- Capability inventory: Uses
curl,node, and environment variable access as part of its operations. - Sanitization: No explicit sanitization or filtering of external content is described in the skill.
Audit Metadata