base

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The skill contains instructions for extracting API keys from a centralized access file located in user documents, which represents credential harvesting behavior.

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill prompt instructs the agent to read and parse a user's centralized access keys file (e.g., ~/Documents/Access.txt) to extract API keys and write them to a project .env file. While reading local files and writing .env files is a common setup pattern, instructing an LLM to read a raw access keys file containing sensitive credentials into its context and process them creates a high risk of secret exposure through the LLM context and conversation logs.