
Sentry Agent Skills
https://github.com/getsentry/sentry-agent-skills- Category
- AI Agents
- Rank
- No. 1657Tools index
Previous survey · No. 1633 ·
- Pricing
- Open Source
- Type
- TOOL
- Builder
- getsentry
- GitHub
- 21 stars
- Latest release
- latest
- Date
About
Collection of agent skills published by Sentry for AI coding assistants to integrate with Sentry workflows.
What it does
It gives coding agents structured playbooks for configuring observability and resolving production errors. The playbooks inspect project dependencies, recommend relevant features, guide implementation, and require verification before completion.
Why it's ranked here
The material is unusually concrete, with platform detection, framework-specific guidance, troubleshooting, and security rules for untrusted event data. However, this repository is archived and explicitly directs new users to its successor.
What's good
Coverage spans Go, Python, Ruby, React, Next.js, .NET, Apple platforms, React Native, and Svelte. Workflows separate discovery, recommendations, implementation, and verification. The issue-fixing guidance also requires source-code corroboration and redaction of secrets and personal data.
Tradeoffs
This is instructional content for compatible coding assistants, not a standalone monitoring product. Some workflows require Sentry MCP, GitHub CLI, authentication tokens, or external documentation checks. The repository receives no further updates, so its embedded SDK guidance can age.
How to use it well
Use the successor repository when adding Sentry through a supported coding assistant. These playbooks best suit teams wanting guided setup, framework-aware configuration, or evidence-led production debugging. They do not replace Sentry itself, project access, credentials, or broader observability tooling.
Technical notes+
README.md marks this repository archived and names getsentry/sentry-for-ai as the replacement. AGENTS.md defines each skill as skills/<skill-name>/SKILL.md with YAML frontmatter and Markdown instructions, targeting the Agent Skills specification. scripts/generate-skill-list.sh extracts name and description fields for release notes. .github/workflows/release.yml packages the skills directory into sentry-agent-skills.tar.gz and publishes a latest-tag release. skills/sentry-fix-issues/SKILL.md depends on Sentry MCP tools, treats event data as untrusted, and requires codebase cross-checking before changes. The provided SDK skills declare license: Apache-2.0 and use phased detection, recommendation, configuration, verification, and troubleshooting guidance.
Observed
- License
- Apache-2.0 is declared in the provided skill frontmatter.
- Primary format
- Markdown instructions with YAML frontmatter.
- Install surface
- Skills CLI, manual directory copying, release tarball, or Sentry wizard.
- Interface
- Agent Skills specification; the issue-fixing workflow additionally requires Sentry MCP.
- Supported clients
- Claude Code, OpenAI Codex, GitHub Copilot, Cursor, OpenCode, and AmpCode.
- Repository status
- Archived and superseded by getsentry/sentry-for-ai.
Read from README.md, AGENTS.md, scripts/generate-skill-list.sh, .github/workflows/release.yml, skills/sentry-go-sdk/SKILL.md, skills/sentry-ruby-sdk/SKILL.md, skills/sentry-cocoa-sdk/SKILL.md, skills/sentry-react-sdk/SKILL.md, skills/sentry-dotnet-sdk/SKILL.md, skills/sentry-fix-issues/SKILL.md, skills/sentry-nextjs-sdk/SKILL.md.
What it can do
Create Sentry issues from detected errors
Error data and stack traces → Sentry issue with metadata
Query Sentry project performance metrics
Project ID and time range → Performance data and statistics
Retrieve Sentry error event details
Event ID or issue ID → Detailed error information and context
Update Sentry issue status and assignments
Issue ID and status/assignment data → Updated issue confirmation
Search Sentry issues by filters
Search criteria and project parameters → Filtered list of matching issues
Configure Sentry project settings
Project configuration parameters → Updated project settings
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.