Sentry for AI
https://github.com/getsentry/sentry-for-ai- Category
- Developer Tools
- Rank
- No. 644Tools index
- Pricing
- Open Source
- Type
- TOOL
- Builder
- getsentry
- GitHub
- 263 stars
- Date
About
Teaches your AI coding assistant how to use Sentry — setup, debugging, alerts — via drop-in skills and configs.
What it does
Sentry for AI is the source library behind assistant-specific Sentry plugins. Its skills guide agents through instrumentation, production issue triage, code fixes, alerts, OpenTelemetry connections, and AI call monitoring. Builds combine shared platform and monitoring references into self-contained packages, while an automatically configured MCP server provides access to Sentry environments.
Why it's ranked here
The case is strong because the project covers both setup and operational debugging, then ships tailored builds for several major coding assistants. Shared references reduce duplicated guidance, and built-artifact link validation catches broken documentation. The main caveat is distribution complexity: this repository is source material, so users must install a generated package elsewhere.
What's good
The workflow reaches beyond initial SDK installation. An agent can identify impactful errors, gather issue context, optionally use Seer analysis, apply fixes, and resolve issues. Instrumentation spans errors, tracing, logs, metrics, profiling, replay, feedback, cron monitoring, and AI calls. Each shipped skill receives its declared references, making the installed artifact self-contained.
Tradeoffs
You cannot install this repository directly. Each supported assistant receives a separately generated plugin, which adds another distribution layer between source changes and users. Some workflows require the GitHub CLI. Installer telemetry starts enabled unless users disable it with a command flag or the standard do-not-track environment setting.
How to use it well
Use it when a team already relies on Sentry and wants its coding assistant to handle setup, incident triage, pull request findings, alerts, or telemetry configuration. Start with the guided entry point, then let platform detection route instrumentation work. It does not replace the Sentry environment, and contributors should work in the source repository rather than editing generated distributions.
Technical notes+
package.json defines a private pnpm workspace. packages/installer/rolldown.config.mjs builds packages/installer/src/index.ts as a CommonJS Node CLI with a shebang while leaving dependencies and Node built-ins external. The CLI exposes install, remove, and uninstall flows, defaults to installation, detects Claude Code, Codex, Cursor, and Grok concurrently, and supports noninteractive selection. packages/installer/src/instrument.ts enables Sentry crash telemetry by default, honors --no-telemetry and DO_NOT_TRACK, and flushes captured failures. scripts/hydrate-references.py copies declared shared references into built skills and removes manifests. scripts/validate-skill-links.py validates relative Markdown links against built artifacts. scripts/gen-semantics.py regenerates checked-in stable semantic-convention references.
Observed
- License
- MIT
- Languages
- TypeScript for the installer, Python for repository scripts, and shell build scripts
- Install surface
- Generated plugins for Agent Plugins, Claude Code, Cursor, Codex, and Grok; skills are also served over HTTP and through skills.sh
- Interface
- Node CLI with install, remove, and uninstall commands, plus automatic Sentry MCP server configuration
- Packaging
- Private pnpm source workspace; the repository itself is not an installable plugin
- Build structure
- Shared reference material is hydrated into each built skill, then relative Markdown links are validated on the built artifact
Read from README.md, package.json, packages/installer/rolldown.config.mjs, packages/installer/src/ui.ts, packages/installer/src/run.ts, packages/installer/src/index.ts, packages/installer/src/system.ts, packages/installer/src/clipboard.ts, packages/installer/src/instrument.ts, packages/installer/src/text-shimmer.ts, packages/installer/src/harnesses/index.ts, scripts/gen-semantics.py, scripts/hydrate-references.py, scripts/validate-skill-links.py, LICENSE.
What it can do
Teach AI assistant Sentry setup procedures
AI coding assistant and Sentry configuration requirements → AI assistant trained with Sentry setup skills
Configure debugging workflows for AI assistant
Debugging scenarios and Sentry debugging tools → AI assistant capable of Sentry debugging tasks
Train AI assistant on Sentry alert management
Alert configuration patterns and Sentry alert system → AI assistant skilled in creating and managing Sentry alerts
Install drop-in Sentry skills to AI assistant
Pre-built Sentry skill modules → AI assistant with integrated Sentry capabilities
Apply Sentry configuration templates
Sentry config templates and target environment → Configured Sentry integration setup
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.