The Agent Skills Directory

[SAFE]: The skill implements a secure workflow for Jira operations, emphasizing user authorization for all mutations (create, edit, delete) and recommending non-interactive, script-friendly modes to avoid unexpected behavior.
[DATA_EXPOSURE]: Instructions explicitly warn against exposing sensitive data like JIRA_API_TOKEN or token values during identity checks (jira me). It encourages using secure storage methods like .netrc or keychains.
[COMMAND_EXECUTION]: The skill uses the jira CLI tool. It provides guidance on safe command construction, such as quoting values and using templates or stdin for large inputs to prevent shell interpolation issues.
[INDIRECT_PROMPT_INJECTION]: As the skill reads content from Jira issues (comments, descriptions), there is a surface for indirect prompt injection. However, the requirement for explicit user authorization for all mutations provides a necessary human-in-the-loop safety control.

using-jira-cli