“observability”的搜索结果

梳理持续运行的 AI 智能体的启停、日志监控、权限管理、分阶段更新与回滚流程，支持故障响应，以及成功率、重试次数、恢复时间和成本评估。

• 设计持续运行智能体的运维流程
• 应对失败次数突然增加
• 制定更新和回滚步骤

依据browser-use参考资料，协助编写Python浏览器自动化代码，涵盖安装、智能体与浏览器配置、模型连接、自定义工具、运行监控及费用跟踪。

• 用Python自动执行浏览器操作
• 配置智能体与浏览器认证
• 添加自定义工具和执行钩子

将应用的LLM调用接入Caveman Cloud网关，在不改变发送内容的情况下计量用量与费用，涵盖连接配置及真实请求验证。

• 开始计量LLM应用的用量与费用
• 更改现有LLM调用的连接地址
• 通过真实请求验证连接

为生产功能设计日志、指标、跨服务追踪和告警，记录请求标识与处理起点，并验证实际输出和通知，帮助了解运行状态及故障原因。

• 完善生产功能的监控信息
• 关联不同服务的日志
• 审查告警与响应操作手册

汇总指定时段已记录的任务完成数、错误数、内存用量和词元消耗等指标，与历史基准比较并标记异常，帮助查看智能体群的运行情况。

• 查看最近一小时的完成数和错误数
• 检查活跃数量、内存和词元消耗
• 查找处理耗时的异常变化

通过Monitor监控由多个智能体组成的swarm，在状态或智能体数量变化时发送通知。支持设置监控间隔，也可单次检查状态或健康状况。

• 接收swarm状态变化通知
• 跟踪智能体数量变化
• 调整监控更新间隔

收集任务中已记录的执行区间，以树状结构整理智能体之间的处理流程，帮助查看耗时、等待时间、错误及可能拖慢任务的环节。

• 调查任务缓慢的原因
• 检查智能体之间的协作
• 定位执行记录中的错误

将已记录的AI使用费用和预算信息导出为Prometheus指标或JSON，用于外部仪表盘展示、预算使用率监控或向Webhook报告。

• 更新Grafana或Prometheus费用指标
• 向Webhook发送一次性费用报告
• 监控预算使用率

Set up or review production error monitoring to capture unhandled errors, failed requests, and performance signals. Check event context, release labels, and alert routing so teams receive useful information when problems occur.

中文简介正在准备中。当前显示英语简介。

• Setting up error monitoring for a new production app
• Investigating errors that monitoring fails to capture
• Reviewing critical alerts and on-call routing

Reconstruct how an AI coding agent produced a feature, pull request, or report using local ax records. Connect sessions, commits, skills, and tool activity into an evidence-based sequence with key decisions, checks, and a repeatable recipe.

中文简介正在准备中。当前显示英语简介。

• Tracing how a shipped feature was built
• Reconstructing development from a commit
• Finding the workflow behind work done on a known date

Deploy and operate AI agents with AWS Bedrock AgentCore. Choose services for hosting agents, connecting external APIs, managing credentials and conversation state, monitoring activity, and assessing agent quality through automated evaluations.

中文简介正在准备中。当前显示英语简介。

• Plan an agent deployment on AgentCore
• Expose a REST API as an agent tool through MCP
• Configure credentials and access controls

Design mock APIs, simulated services that reproduce expected responses, authentication, errors, and delays. Define scenarios and repeatable test data to support frontend development, integration testing, and realistic demos before the real backend is ready.

中文简介正在准备中。当前显示英语简介。

• Developing a frontend before the backend is ready
• Simulating a partner or third-party API
• Preparing a demo with realistic API responses

Coordinate specialist agents to investigate and improve application performance across databases, backend services, frontend code, and delivery infrastructure. Establish baseline measurements, address observed bottlenecks, validate changes with load tests, and set up monitoring to detect future slowdowns.

中文简介正在准备中。当前显示英语简介。

• Finding bottlenecks across an entire application
• Coordinating database, backend, and frontend improvements
• Validating performance changes with load tests

Inspect local AI coding session logs with agenttrace to identify cost spikes, delays, failed tools, and retry loops. Produce readable reports, compare attempts, and configure automated checks for unusual or unhealthy runs.

中文简介正在准备中。当前显示英语简介。

• Investigate slow or expensive AI coding sessions
• Review failed runs before retrying
• Compare tool usage and verification across attempts

Add monitoring to AI agents with response times, token usage, estimated costs, and error metrics. Trace records follow model and tool calls step by step, helping teams investigate failures and understand operating costs.

中文简介正在准备中。当前显示英语简介。

• Investigate slow agent responses
• Track token usage and estimated costs
• Trace failures in model and tool calls

Set up distributed tracing with Jaeger and Tempo to follow requests across services. Examine delays, service dependencies, and error propagation, with guidance on sampling, linking logs to traces, and troubleshooting missing trace data.

中文简介正在准备中。当前显示英语简介。

• Investigating slow requests across services
• Understanding service dependencies
• Tracing how errors spread

Build Grafana dashboards that show application and infrastructure health through request rates, errors, response times, and resource usage. Organize metrics into useful panels, add filters and alerts, and manage dashboard definitions as code.

中文简介正在准备中。当前显示英语简介。

• Build an API health dashboard
• Visualize CPU and memory usage
• Track service level objectives and business KPIs

Set up Phoenix tracing for LLM applications in Python or TypeScript. The skill guides automatic and custom recording of individual operations, conversation grouping, metadata, feedback, and production settings such as batching and personal data masking.

中文简介正在准备中。当前显示英语简介。

• Set up tracing for an LLM application
• Record document retrieval and tool calls
• Group conversation traces by session

Add Arize AX tracing to an LLM application, recording how model calls and tool executions connect. Inspect the existing code, select suitable integrations, add instrumentation, and verify that the application runs and traces reach Arize.

中文简介正在准备中。当前显示英语简介。

• Setting up tracing for an LLM application
• Recording tool inputs and outputs in an agent
• Adding Arize export to an existing OpenTelemetry setup

Apply consistent names to SGLang speculative decoding code, which verifies draft tokens to generate text. Distinguish tokens with or without the bonus token, counts, counters, and arrays when adding identifiers, renaming code, or reviewing changes.

中文简介正在准备中。当前显示英语简介。

• Naming variables in speculative decoding code
• Renaming identifiers to match their meaning
• Reviewing metric and counter names

Design LLM applications using patterns for answers grounded in documents, agents that use tools, and prompt testing. Implementation examples also cover monitoring response quality, latency, costs, and failures during operation.

中文简介正在准备中。当前显示英语简介。

• Building an assistant that answers from internal documents
• Choosing an architecture for an agent that uses tools
• Comparing and versioning prompts

Investigate production issues by searching Datadog logs, querying metrics, and following request traces across services. Compare error counts across time periods, inspect surrounding logs, and manage dashboards through a command-line tool.

中文简介正在准备中。当前显示英语简介。

• Investigating errors in production
• Comparing error counts across time periods
• Following request traces across services

Connect cognee to additional databases and data sources through community packages. The skill covers installing and registering adapters, importing data through connectors, configuring monitoring, and developing extension packages with examples and tests.

中文简介正在准备中。当前显示英语简介。

• Connect cognee to Qdrant or another database
• Import data from Slack or Gmail
• Configure monitoring with Keywords AI

Audit a codebase before production deployment across eight areas, including security, databases, dependencies, and AI usage. Receive findings with file locations, manual checklists, and a release verdict that blocks shipping when critical issues remain.

中文简介正在准备中。当前显示英语简介。

• Check an application before its first production release
• Find security and database issues before deployment
• Review backup restoration and rollback readiness