aiagent.club
English
GitHub

parlant

观察池 · 暂无正式排名

该项目当前未满足“两个有效维度 + 两种数据源”的主榜门槛。

未排名
可观测采用度缺失
动量当前有效 · 2026-09-12
12
关注度当前有效 · 2026-09-12
45
信号可信度 依据当前有数据的独立评分维度数量计算。
2/3 · 1 种数据源

方法论 v2.0 · 快照 2026-09-12 · 超过 2 天视为过期

项目介绍

Website • Quick Start • Examples • Discord

Looking for an open-source alternative to Ada, Decagon, or Sierra?

Parlant is production-ready. It streamlines the development and maintenance of enterprise-grade B2C (business-to-consumer) and sensitive B2B interactions that need to be consistent, compliant, on-brand, and comprehensively traceable.

Conversational context engineering is hard because real-world interactions are diverse, nuanced, and non-linear.

System prompts work until production complexity kicks in. The more instructions you add to a prompt, the faster your agent stops paying attention to any of them.

Routed graphs solve the prompt-overload problem, but the…

各数据源

18.3k Star
  • Star 18.3k
  • Fork 1.6k
  • 提交 5.5k
  • 发布 33