Install any skill in seconds. Free to start, no credit card required.
Get Started Free →tsinghua-fib-lab/social-media
社交媒体互动:发帖 / 转发 / 评论 / 点赞 / 关注、刷新 Feed 推荐流、搜索话题贴文。必须通过 ask_env 调用 SocialMediaSpace 工具。
Passed 8 of 15 test cases with the skill, and 2 of 15 without it.
and 16 more in the full report
10 fixed3 passed either way10 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.
As user ID 7, I want to unfollow user ID 3. Should I use unfollow_user with user_id and target_id as keys?
| Checked | Without skill | With skill |
|---|---|---|
| Uses instruction template unfollow_user follower_id={user_id} followee_id={followee_id}. | Missed | Passed |
| Sets user_id to 7 and followee_id to 3 in variables. | Missed | Passed |
The instructions your model reads when the skill loads.
SocialMediaSpace 是微博 / Twitter 风格的社交媒体环境。所有操作通过 ask_env 调用,环境路由器会自动将指令转换为 SocialMediaSpace 工具调用。
默认 agent_id === user_id:你在社交媒体中的 id 就是你的 agent id。ctx={"id": <your_id>} 提供你的身份;凡 author_id / user_id / follower_id 都填自己的 id,只有对他人发起动作(关注)时才用对方 id。
每条帖子有 post_id、author_id、content、tags(话题标签列表)、post_type(original / repost / comment),以及 likes_count / comments_count / reposts_count / view_count。
refresh_feed 返回按算法排序的贴文流:chronological(时间倒序,默认)、reddit_hot(热度)、twitter_ranking(综合社交关系)、random、mf(预训练模型)。这是贴文流推荐(Timeline),不是电商物品推荐。
所有操作通过 ask_env 调用,用模板指令 + variables 让同类调用结构一致(易变值放 variables,指令措辞保持稳定):
ask_env(instruction="<指令模板>", variables={...}, ctx={"id": <your_id>}, readonly=<bool>)
观察自己的状态:资料(粉丝 / 关注 / 发帖数)、最近 Feed(5 条)、收到的互动、近期活动、社交关系变动、可用行为列表。每个 step 先观察。
ask_env(instruction="<observe>", ctx={"id": <your_id>})
发布原创帖子。tags 为话题标签列表。
ask_env(
test cases passed, vs 2 / 15 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.