The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the gh (GitHub CLI) to manage issue trackers. The scripts/frontier.py script invokes shell commands using list-based arguments and validated inputs (e.g., integer issue numbers), which prevents command injection vulnerabilities.
[CREDENTIALS_SAFE]: The skill implements a concurrency mechanism for ticket claims. Local lock files are created using os.O_EXCL for atomicity and 0o600 permissions, ensuring that session identifiers are only accessible by the owner, following the principle of least privilege.
[SAFE]: The skill relies on the official GitHub CLI and standard Python libraries. No external package downloads or remote code execution patterns from untrusted sources were identified.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from GitHub issues and comments. While this presents a surface for indirect prompt injection, the skill's workflow is strictly limited to planning and metadata management, with no capabilities for implementation or execution of code derived from these inputs.

decision-map