The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of spawnSync to execute git, gh (GitHub CLI), and curl commands. These executions are core to the skill's functionality, such as checking repository status, managing worktrees, and interacting with issue trackers. The skill follows the practice of using array-based arguments for subprocess calls, which helps prevent shell injection vulnerabilities.- [EXTERNAL_DOWNLOADS]: In scripts/issue-media.mjs, the skill implements logic to download images referenced in issue metadata using curl. A security function isTrustedAuthUrl is implemented to ensure that authentication headers (like Bearer or Basic tokens) are only sent to trusted hostnames (e.g., github.com or the configured Jira/Confluence instance), effectively preventing credential leakage to arbitrary remote servers.- [DATA_EXFILTRATION]: The skill transmits issue data and documentation reports to external services (Jira and Confluence). This is the primary intended functionality of the skill. Sensitive tokens are retrieved from environment variables defined in the user's configuration (e.g., JIRA_API_TOKEN), which is a standard and secure method for handling API authentication.- [PROMPT_INJECTION]: The skill processes untrusted external data, such as issue titles and descriptions from GitHub or Jira. While this provides a potential surface for indirect prompt injection, the skill mitigates this by enforcing human-in-the-loop approval (AskUserQuestion) before any content is submitted back to the tracker or acted upon by the agent.- [SAFE]: The code is well-structured, documented, and free of obfuscation. It manages its own workspace within the .issue/ directory and correctly updates the project's .gitignore to ensure that local metadata and temporary files are not accidentally committed to the repository.

issue-create