The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill requires the user to set a sensitive blockchain private key in the FITTINGS_PRIVATE_KEY environment variable. The documentation notes that this key sits in plaintext where the agent runs and recommends using a throwaway wallet.
[EXTERNAL_DOWNLOADS]: The documentation provides instructions to download the execution script using curl -sO https://fittings.sh/skill/call.mjs if it is missing from the local environment.
[REMOTE_CODE_EXECUTION]: The call.mjs script dynamically imports several Node.js packages at runtime, including @x402/fetch, @x402/evm, and viem. Additionally, the skill's primary function involves executing a local script (node call.mjs) which is also served from a remote URL.
[COMMAND_EXECUTION]: The skill instructions direct the agent to execute shell commands (node call.mjs, export, curl) to perform its primary functions.
[DATA_EXFILTRATION]: The call.mjs script uses the provided FITTINGS_PRIVATE_KEY to sign payment authorizations that are transmitted to the fittings.sh gateway to settle service fees.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external websites (via the domain input) through the fittings.sh proxy to extract logos.
Ingestion points: The domain parameter triggers a remote fetch of a website's homepage contents via a proxy service.
Boundary markers: None identified in the script or instructions to separate external content from agent instructions.
Capability inventory: The script has capabilities for network access (fetch), file system reads (readFileSync), and shell execution.
Sanitization: There is no evidence of local sanitization of the data returned from the remote service before it is presented to the agent context.

brand-logo