The Agent Skills Directory

[PROMPT_INJECTION]: The skill facilitates an indirect prompt injection surface by interpolating untrusted user data into instructions for spawned sub-agents.
Ingestion points: User-supplied variables such as {ISSUE_DESCRIPTION}, {FEATURE_DESCRIPTION}, and {MODULE_DESCRIPTION} are read in SKILL.md and passed to the Task tool.
Boundary markers: There are no explicit delimiters or instructions provided to sub-agents to disregard instructions embedded within the user-provided variables.
Capability inventory: The spawned sub-agents have the capability to modify the local filesystem and execute code. The lead agent also possesses command execution capabilities for Magento maintenance.
Sanitization: No input validation, escaping, or filtering of user-provided strings is performed before they are incorporated into the sub-agent spawn prompts.
[COMMAND_EXECUTION]: The skill executes bin/magento commands (setup:di:compile and cache:flush) during the final review phase. These commands are standard for the Magento 2 development workflow and are used as intended for project maintenance.

agent-teams