The Agent Skills Directory

[DATA_EXPOSURE]: The skill documentation and code examples correctly emphasize the use of environment variables and KMS for managing sensitive information like SECRET_KEY and MYSQL_PASSWORD. The templates provided in the documentation use process.env rather than hardcoding credentials.
[COMMAND_EXECUTION]: The skill focuses on Node.js 18 runtime environments. Analysis of the provided code examples in SKILL.md and EXAMPLES.md shows no usage of dangerous system command execution functions like child_process.exec or eval.
[EXTERNAL_DOWNLOADS]: The skill references standard, well-known libraries such as tencentcloud-sdk-nodejs, cos-nodejs-sdk-v5, axios, mysql2, and pino. These are legitimate dependencies for the stated purpose of cloud function development.
[PROMPT_INJECTION]: The instructions and rules are strictly professional and technical. There are no attempts to bypass safety filters or override the agent's core instructions.
[DATA_EXFILTRATION]: Network operations are restricted to standard cloud service interactions (Tencent Cloud STS/COS) and database connections. No unauthorized data exfiltration patterns were detected.

scf-worker