The Agent Skills Directory

[SAFE]: No security issues were identified in this skill. It functions as a set of logical instructions for agent behavior during task failure.- [NO_CODE]: The skill is composed purely of markdown text and does not include scripts, binaries, or external dependencies.- [PROMPT_INJECTION]: The use of strong instructional language (e.g., 'HARD-GATE', 'MUST', 'RIGID') serves to define internal logic for retrying failed tasks and does not attempt to subvert platform safety constraints, bypass filters, or override primary system prompts.

resilient-execution