Sentry Toolkit provides agentic tooling and Model Context Protocol integrations designed specifically for human-in-the-loop coding assistants such as Cursor and Claude Code. Acting as middleware to the Sentry API, it enables AI agents to inspect stack traces, query logs, and triage issues using natural language translation powered by embedded LLMs. The toolkit supports both remote cloud transports and local stdio execution with granular skill controls, allowing developers to diagnose and resolve production exceptions seamlessly within their development workflow.
主要功能
01Natural language AI search across Sentry errors, traces, logs, and issues
02Built-in rate limiting resilience and automated transient request retries
03949 GitHub stars
04Dual transport options supporting both remote HTTP/OAuth and local stdio modes
05Claude Code plugin integration for automatic subagent delegation
06Configurable skill sets with full support for self-hosted Sentry instances
使用案例
01Investigating and resolving production errors directly from within AI code editors
02Querying complex Sentry telemetry using plain English natural language requests
03Delegating issue triage and root-cause analysis to autonomous development agents