Skills matrix
| Skill | Status | Delivery | Maturity | Evidence |
|---|---|---|---|---|
| Code generation | Partial | Model-dependent | Model-limited | Not publicly documented |
| Refactoring | Partial | Model-dependent | Model-limited | Not publicly documented |
| Multi-file reasoning | Partial | Model-dependent | Model-limited | Not publicly documented |
| Test generation | Partial | Model-dependent | Model-limited | Not publicly documented |
| Static analysis support | Partial | Model-dependent | Model-limited | Not publicly documented |
| Codebase indexing | Partial | Model-dependent | Model-limited | Not publicly documented |
| Semantic retrieval | Partial | Model-dependent | Model-limited | Not publicly documented |
| Memory retention across sessions | Partial | Model-dependent | Model-limited | Not publicly documented |
| Context injection control | Partial | Model-dependent | Model-limited | Not publicly documented |
| Inline editing | Partial | Model-dependent | Model-limited | Not publicly documented |
| Chat-first interaction | Partial | Model-dependent | Model-limited | Not publicly documented |
| File diff preview | Partial | Model-dependent | Model-limited | Not publicly documented |
| Git integration | Partial | Model-dependent | Model-limited | Not publicly documented |
| Terminal / command execution | Partial | Model-dependent | Model-limited | Not publicly documented |
| Model switching | Partial | Model-dependent | Model-limited | Not publicly documented |
| Multi-model orchestration | Partial | Model-dependent | Model-limited | Not publicly documented |
| Prompt augmentation layer | Partial | Model-dependent | Model-limited | Not publicly documented |
| Agent loop execution | Partial | Model-dependent | Model-limited | Not publicly documented |
| Local model support | Partial | Model-dependent | Model-limited | Not publicly documented |
| Offline capability | Partial | Model-dependent | Model-limited | Not publicly documented |
| Enterprise policy control | Partial | Model-dependent | Model-limited | Not publicly documented |
Why it matters
- Python-specialized analysis understands idiomatic Python — it knows when to use list comprehensions, context managers, dataclasses, and other Pythonic patterns.
- Automated PR reviews catch code quality issues before human reviewers spend time on style and structure — freeing reviews for logic and architecture.
- IDE integration (VS Code, PyCharm) delivers suggestions in-editor, not just on PRs — shifting quality enforcement left in the development cycle.
- Explains every suggestion with reasoning, making it a learning tool for junior developers improving Python skills.
Key capabilities
- Automated PR review: Comments on GitHub/GitLab PRs with refactoring suggestions and quality improvements.
- Refactoring suggestions: Extract methods, simplify conditionals, convert loops to comprehensions, remove duplication.
- Anti-pattern detection: Flag Python-specific issues: mutable default arguments, bare except clauses,
==vsisfor None. - IDE extension: VS Code and PyCharm plugins for real-time suggestions as you code.
- Metrics dashboard: Track code quality trends, review coverage, and most common issues across the codebase.
- Custom rules: Define team-specific coding standards and enforce them in PR reviews.
- GitHub Actions integration: Run Sourcery as part of CI/CD pipeline checks.
- Batch refactoring: Apply suggested changes to the entire codebase via CLI.
Technical notes
- Language focus: Python (primary); some support for JavaScript/TypeScript
- Integrations: GitHub, GitLab, Bitbucket (PR review); VS Code, PyCharm (IDE)
- CI/CD: GitHub Actions, GitLab CI integration
- CLI:
sourcery reviewfor batch analysis;sourcery loginfor auth - Pricing: Free (open source); Team ~$20/user/mo; Enterprise custom
- Founded: 2019; London, UK; backed by Accel, LocalGlobe
Ideal for
- Python engineering teams who want automated code quality enforcement on every PR without a dedicated code review tool.
- Teams with junior developers where consistent Pythonic style and refactoring feedback accelerates learning.
- Organizations with existing Python codebases who want to reduce technical debt through systematic refactoring suggestions.
Not ideal for
- Polyglot teams needing review across many languages — CodeRabbit or DeepCode cover more languages.
- Security-focused review — Sourcery catches quality issues, not security vulnerabilities (use Snyk or Semgrep for security).
- Teams working primarily in JavaScript, Go, Java, or other non-Python languages.
See also
- CodeRabbit — Language-agnostic AI PR review with broader language support.
- What the Diff — AI-generated PR summaries for non-technical stakeholders.
- Aider — AI pair programmer that edits code directly, not just reviews it.