The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the vefaas command-line interface through the Bash(vefaas:*) tool. This is the primary mechanism for managing cloud resources such as functions, applications, and sandboxes.
[EXTERNAL_DOWNLOADS]: The skill provides instructions for installing and updating the vendor-provided CLI tool using npm i -g @volcengine/vefaas-cli@latest. This is a standard procedure for a well-known service provided by the skill's authoring organization.
[DATA_EXFILTRATION]: The skill explicitly instructs the agent not to display sensitive information like Access Keys, Secret Keys, or session tokens to the user. It emphasizes using environment variables or the system keychain for secure credential storage.
[PROMPT_INJECTION]: While the skill uses strong markers like 'CRITICAL' and 'MUST' in its instructions, these are employed for operational workflow control (e.g., version checks and intent disambiguation) rather than attempting to override the AI's core safety guidelines or bypass system constraints.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface by reading logs and project source code. However, it implements a 'High Risk Operation Protocol' requiring explicit user confirmation before executing destructive or state-changing commands, which serves as a mitigation for potential injection from untrusted inputs.
[REMOTE_CODE_EXECUTION]: No patterns of remote script execution (e.g., curl to bash) or arbitrary dynamic code generation from untrusted sources were detected.

byted-vefaas-skills