The Agent Skills Directory

[SAFE]: The skill implements a logical classification system for project tickets. It calculates complexity scores based on the presence of technical keywords and labels to guide the agent's workflow depth.
[COMMAND_EXECUTION]: While the skill mentions integration with commands like /resolve and /analyze-ticket, and recommends launching exploration agents or the 'Architect' skill, these are standard workflow orchestration patterns within the agent environment and do not involve arbitrary shell execution or system modification.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (ticket titles, descriptions, and comments). There is a minor risk that an attacker could attempt to manipulate the complexity score by stuffing a ticket with specific keywords (e.g., repeatedly using 'quick-fix' to force a SIMPLE classification). However, since the output is used for internal workflow prioritization and doesn't trigger sensitive operations, the impact is negligible.
[DATA_EXPOSURE]: The skill reads configuration from .claude/ticket-config.json. This is a standard practice for skill configuration and does not represent an exposure of sensitive user or system data.

analyze-ticket