The Agent Skills Directory

[SAFE]: No security issues or malicious patterns were identified. The skill implements standard session management and orchestration logic.- [COMMAND_EXECUTION]: The skill uses shell commands through the Bash tool to interact with project-local utilities like git and the GitHub CLI (gh). These operations are restricted to the intended purpose of managing commits and issues within the current repository.- [DATA_EXFILTRATION]: Progress tracking is performed using a local state file (.claude/pipeline-state.json). No unauthorized network activity or transmission of sensitive project data to external domains was detected.- [EXTERNAL_DOWNLOADS]: The skill does not download or execute remote scripts or external dependencies at runtime. It relies on standard system utilities like jq for status reporting.

checkpoint-resume