PrismCat

中文

v1.7.0 让 PrismCat 在可观测性与用量统计上有了显著增强，新增 Trace 关联与详情页、Usage 提取规则库、Body 关键词搜索，以及更合理的大 Body 存储策略。

• 新增 Trace 关联能力以及 Traces / Trace Detail 页面，可从完整调用链回看多条日志
• 新增 Usage / token 用量提取能力，支持按上游绑定规则并在设置页管理规则库
• 内置 Usage 模版扩展为 OpenAI compatible、OpenAI Responses、Anthropic、Gemini，适配 Responses API、Codex 等新工作流
• 请求 / 响应 Body 查看器新增关键词搜索能力
• 优化大 Body 的存储与展示，数据库保留可读预览，完整内容按分离存储处理
• Usage 提取结果已接入日志详情与汇总展示
• 持续优化日志详情页的文本展示与阅读体验
• 改进格式化 Body 与分离存储 Body 的处理方式
• 手动二进制发布 workflow 的默认标签已更新为 v1.7.0
• 注意：本版本调整了 Body 日志存储格式，旧版本生成的大 Body 日志升级后可能无法正确加载分离内容，如有重要旧日志请提前备份或导出

English

v1.7.0 brings major improvements to observability and usage tracking in PrismCat, adding Trace correlation with dedicated pages, a Usage extraction rule library, Body keyword search, and a smarter storage strategy for large bodies.

• Added Trace correlation along with Traces / Trace Detail pages for reviewing multiple logs across a full call chain
• Added Usage / token extraction with per-upstream binding rules, managed via a rule library on the settings page
• Expanded built-in Usage templates to OpenAI compatible, OpenAI Responses, Anthropic, and Gemini, covering Responses API, Codex, and other newer workflows
• Added keyword search to the request / response Body viewer
• Improved storage and display of large Bodies, keeping a readable preview in the database while storing full content separately
• Wired Usage extraction results into log detail and summary views
• Continued refinement of text rendering and readability on the log detail page
• Improved handling of formatted Bodies and separately-stored Bodies
• Updated the manual binary release workflow default tag to v1.7.0
• Note: this version changes the Body log storage format. Large-Body logs produced by older versions may fail to load their detached full content after upgrading — please back up or export important historical logs beforehand

Full Changelog: https://github.com/paopaoandlingyia/PrismCat/compare/v1.6.0...v1.7.0