Issues

• enhancement: 优化向大模型发送对话历史时的 Token 消耗

  enhancement
  agent-topia/evolving_personality#48 · Filter by author Xingzhi107 opened on Mar 26, 2026

• bug: 记忆状态更新时，偶尔会导致前文语境（Context）断层。

  bug
  agent-topia/evolving_personality#47 · Filter by author ZHONG080901 opened on Mar 26, 2026

• Integration Idea: OpenClaw Skill Integration

  agent-topia/evolving_personality#46 · Filter by author baofuwu opened on Mar 26, 2026

• feature: 设计一套跨模型的统一异常处理（Exception Handling）中间件

  feature
  agent-topia/evolving_personality#45 · Filter by author FallenYing opened on Mar 26, 2026

• bug: 当使用较弱的模型（如Llama-7B早期版本）时，Reflection 生成的摘要逻辑混乱。

  bug
  agent-topia/evolving_personality#44 · Filter by author jerry-hs opened on Mar 26, 2026

• bug: 并发请求多个 Qwen Agent 实例时，存在速率限制（Rate Limit）导致崩溃。

  bug
  agent-topia/evolving_personality#43 · Filter by author leezy18 opened on Mar 26, 2026

• enhancement:将测试框架与 Weights & Biases (WandB) 集成，方便打点和实验追踪

  enhancement
  agent-topia/evolving_personality#42 · Filter by author junmingvuvu-pixel opened on Mar 26, 2026

• feature: 为API请求添加自动重试与指数退避（Exponential Backoff）机制。

  feature
  agent-topia/evolving_personality#41 · Filter by author Chauncey-Hsu opened on Mar 26, 2026

• 自动化计算生成文本的多样性（Diversity）和连贯性（Coherence）指标。

  agent-topia/evolving_personality#40 · Filter by author yunyancuo opened on Mar 26, 2026

• bug: 测试报告生成的 Markdown 表格在某些渲染器下格式错乱。

  bug
  agent-topia/evolving_personality#39 · Filter by author bry-zczh opened on Mar 26, 2026

• [Bug] Llama 模型的 Type Activation Accuracy 在特定场景下波动过大，需校准系统提示词。

  bug
  agent-topia/evolving_personality#38 · Filter by author bread-ovO opened on Mar 26, 2026

• enhancement: 将API调用层重构为工厂模式，方便未来随时接入新大模型。

  enhancement
  agent-topia/evolving_personality#37 · Filter by author Arui5566 opened on Mar 26, 2026

• feature: 扩充论文提到的“类型特定场景测试”（Type-specific scenario testing）数据集

  feature
  agent-topia/evolving_personality#36 · Filter by author tbjuechen opened on Mar 26, 2026

• 添加 HuggingFace transformers 原生本地加载与推理的脚本。

  agent-topia/evolving_personality#35 · Filter by author lvxingye opened on Mar 26, 2026
  1

• feature: 引入第三方人类专家评分（Human Eval）的数据记录与清洗脚本

  feature
  agent-topia/evolving_personality#34 · Filter by author KimmyXYC opened on Mar 26, 2026

• feature: 支持从JSON/YAML恢复并热加载“人格快照”

  feature
  agent-topia/evolving_personality#33 · Filter by author AsakiRain opened on Mar 26, 2026

• feature: 增加 Azure OpenAI 服务的 API 端点和鉴权支持

  feature
  agent-topia/evolving_personality#32 · Filter by author RC-CHN opened on Mar 26, 2026

• enhancement:

  enhancement
  agent-topia/evolving_personality#31 · Filter by author shiyizhiyan opened on Mar 26, 2026

• feature: 提供API用于在对话运行中实时修改Agent的环境压力值（Contextual Stress）

  feature
  agent-topia/evolving_personality#30 · Filter by author wsa619 opened on Mar 26, 2026

• feature: 为不同MBTI类型设计专属的Prompt注入模板，以提高角色扮演的深度。

  feature
  agent-topia/evolving_personality#29 · Filter by author yhsyhsyhs7 opened on Mar 20, 2026

• feature: 支持多Agent交互下的人格演化传染机制（如群体极端化或趋同）。

  feature
  agent-topia/evolving_personality#28 · Filter by author chihaoyou opened on Mar 20, 2026

• feature: 改进Reflection机制，使其能够区分“长期性格固化”与“短期情绪波动”。

  feature
  agent-topia/evolving_personality#26 · Filter by author kaisacloud opened on Mar 20, 2026

• feature: 接入向量数据库（如 ChromaDB 或 Qdrant）用于长文本人格记忆的检索

  feature
  agent-topia/evolving_personality#25 · Filter by author Mandaluoren opened on Mar 19, 2026

• feature: 接入向量数据库（如 ChromaDB 或 Qdrant）用于长文本人格记忆的检索

  feature
  agent-topia/evolving_personality#24 · Filter by author Mandaluoren opened on Mar 19, 2026

• feature: 提供 CLI (命令行工具) 用于快速启动一次人格演化对话。

  feature
  agent-topia/evolving_personality#22 · Filter by author lislie422 opened on Mar 19, 2026