Reply to a recent direct or group message without inventing unrelated movement.
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
In `scripts/run_skill.py`, the required workflow reads outsider-authored free text from `observation["recent_messages"]` via `recent_external_messages()`, then injects `last["content"]` into the reply and `memory_template` (`build_reply()` and `memory_effect(SKILL_ID, ...)`).
XiaoLuoLYG/GOD
753f067
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.