• Joined on 2026-06-26
solahqb created branch feat/go-no-go-verdict in solahqb/AgentEvalTool 2026-08-25 06:29:40 +00:00
solahqb pushed to feat/go-no-go-verdict at solahqb/AgentEvalTool 2026-08-25 06:29:40 +00:00
2314ebe3bc feat(report): 上线验收报告 go/no-go 结论
5827c3d3f5 docs(research): go/no-go 验收报告技术方案调研
Compare 2 commits »
solahqb closed issue solahqb/AgentEvalTool#20 2026-08-25 06:23:31 +00:00
上线验收报告 go/no-go 结论调研
solahqb commented on issue solahqb/AgentEvalTool#20 2026-08-25 06:23:30 +00:00
上线验收报告 go/no-go 结论调研

Resolution

调研完成,详见 research/go-no-go-report.md(分支 `research/go-no…

solahqb closed issue solahqb/AgentEvalTool#19 2026-08-25 06:19:34 +00:00
扩展 llm_score 支持多维度评分
solahqb commented on issue solahqb/AgentEvalTool#19 2026-08-25 06:19:34 +00:00
扩展 llm_score 支持多维度评分

Resolution

实现完成,PR #27。

实现内容:

  • RuleResult 新增 details 字段存储结构化多维度分数 -…
solahqb created pull request solahqb/AgentEvalTool#27 2026-08-25 06:19:12 +00:00
feat(llm_score): 支持多维度独立评分
solahqb pushed to feat/llm-score-multi-dimension at solahqb/AgentEvalTool 2026-08-25 06:18:43 +00:00
a29f78ff4b feat(llm_score): 支持多维度独立评分
solahqb created branch feat/llm-score-multi-dimension in solahqb/AgentEvalTool 2026-08-25 06:18:42 +00:00
solahqb closed issue solahqb/AgentEvalTool#18 2026-08-25 06:12:07 +00:00
功能完整性评估方案调研
solahqb commented on issue solahqb/AgentEvalTool#18 2026-08-25 06:12:06 +00:00
功能完整性评估方案调研

Resolution

调研完成,详见 [research/functional-completeness-assessment.md](https://git.solahqb22.cn/solahqb/AgentEvalTool/blob/research/functional-completeness/research/functional-comple…

solahqb created branch research/functional-completeness in solahqb/AgentEvalTool 2026-08-25 06:11:57 +00:00
solahqb pushed to research/functional-completeness at solahqb/AgentEvalTool 2026-08-25 06:11:57 +00:00
51d97694dc research: 功能完整性评估方案调研
01af6fca74 docs(research): 功能完整性评估方案调研
Compare 2 commits »
solahqb opened issue solahqb/AgentEvalTool#26 2026-08-25 05:56:11 +00:00
实现用户体验指标(放弃率+流畅度)
solahqb opened issue solahqb/AgentEvalTool#25 2026-08-25 05:56:11 +00:00
实现成本效率跟踪(对话级+任务级)
solahqb opened issue solahqb/AgentEvalTool#21 2026-08-25 05:55:57 +00:00
扩展 response_time 支持首字延迟和吞吐量
solahqb opened issue solahqb/AgentEvalTool#22 2026-08-25 05:55:57 +00:00
扩展 safety 规则覆盖幻觉、越权和合规
solahqb opened issue solahqb/AgentEvalTool#23 2026-08-25 05:55:57 +00:00
质量监控仪表盘历史趋势调研
solahqb opened issue solahqb/AgentEvalTool#24 2026-08-25 05:55:57 +00:00
能力对比报告雷达图和排名调研
solahqb opened issue solahqb/AgentEvalTool#19 2026-08-25 05:55:31 +00:00
扩展 llm_score 支持多维度评分