The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill explicitly includes functionality for "SSH key generation for SFTP" as noted in the description and references (references/ssh_keys.md). The generation and management of cryptographic private keys by an AI agent presents a significant risk of exposing sensitive credentials if the generated material is displayed in output, stored insecurely, or handled without proper encryption protocols.
[PROMPT_INJECTION]: The skill's core functionality involves processing untrusted external data, which creates a surface for indirect prompt injection attacks.
Ingestion points: The skill is instructed to analyze, validate, and debug external FiveM resources, including Lua scripts, fxmanifest.lua files, and server logs.
Boundary markers: There are no instructions or patterns provided to encapsulate untrusted input within delimiters or to instruct the agent to ignore instructions embedded in the processed code or logs.
Capability inventory: The skill possesses powerful capabilities including script refactoring, database schema optimization (oxmysql), and SSH credential management.
Sanitization: The skill documentation does not outline any sanitization, validation, or filtering steps to be performed on external content before it is processed by the agent.

fivem