The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration attempts were detected in the provided files.
[COMMAND_EXECUTION]: The skill relies on the run_command tool to execute git worktree and GitHub CLI (gh) operations. This is a legitimate and expected use of command execution for a developer utility focused on PR orchestration.
[PROMPT_INJECTION]: Documentation in SKILL.md contains routing directives and usage guidelines intended to help the agent select the correct tool for specific tasks. These instructions are benign and do not attempt to bypass safety filters or override system prompts.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied or external data such as issue titles and descriptions, which constitutes a potential attack surface for indirect prompt injection.
Ingestion points: The issue_pr_orchestrate.py script accepts title and body arguments which may originate from external issue trackers.
Boundary markers: No explicit boundary markers or delimiters for untrusted content are defined in the instructions.
Capability inventory: The skill uses high-privilege tools including run_command, write_to_file, and replace_file_content.
Sanitization: The provided files do not include the implementation details of the orchestration script to verify how inputs are escaped or validated.

issue-pr-lifecycle-agent