Repo Pulse
Aider hits 43k stars amid import errors, Sonnet 4.5 support
Aider has reached 43,644 stars and is actively shipping Claude Sonnet 4.5 support with overeager mode enabled. Recent activity shows heavy issue volume (61 opened, 39 closed in 30 days) dominated by import and runtime errors, with minimal commit velocity.
Aider is an open-source terminal-based AI pair programmer that integrates large language models into local development workflows. The project has accumulated 43,644 GitHub stars and recently shipped support for Claude Sonnet 4.5 with overeager mode enabled. Despite the milestone, recent activity reveals a project managing significant technical debt and a high volume of user-facing errors.
By the numbers
| Metric | Value |
|---|---|
| Current stars | 43,644 |
| 30-day commits | 1 |
| 30-day contributors | 1 |
| 30-day PRs merged | 0 |
| 30-day issues opened | 61 |
| 30-day issues closed | 39 |
| Recent releases (90d) | None |
| Default branch | main |
The activity snapshot reveals a project in maintenance mode. A single commit from paul-gauthier on April 9, 2026, represents the entire 30-day commit history, and no pull requests were merged in that window. This contrasts sharply with the 61 issues opened, suggesting the project is receiving substantial user feedback but shipping minimal code.
What’s shipping
The sole recent commit enables overeager mode for Claude Sonnet 4.5 models. This is a configuration change rather than a structural feature, indicating Aider is optimizing its integration with Anthropic’s latest model release. The “overeager” mode likely adjusts how aggressively the model proposes code changes, a tuning parameter relevant to pair-programming workflows where false confidence can degrade user experience.
No pull requests were merged in the past 30 days, and no releases were cut in the past 90 days. This absence of formal release cadence is notable for a project with 43k stars and active user engagement. Users may be running from the main branch directly or relying on older stable releases, creating a potential gap between what’s deployed and what’s in development.
Open questions
Recent issues cluster around import and runtime errors rather than feature requests. The most visible pattern is uncaught exceptions in utility and type-definition modules:
- Issue 5055: ImportError in _legacy_response.py line 27 (open)
- Issue 5051: ImportError in _types.py line 36 (open)
- Issue 5056: ImportError in utils.py line 43 (closed)
- Issue 5053: RuntimeError in pathlib.py line 1237 (closed)
- Issue 5044: OSError in util.py line 1054 (closed)
Three generic “Bug report” issues (5048, 5047, 5046) opened within seconds of each other on April 18 suggest either automated testing or a batch of users encountering similar failures simultaneously. The rapid closure of some import errors (5056, 5044) indicates the team is responsive to critical failures, but the open backlog of import issues suggests either incomplete fixes or a new regression.
The import errors are particularly concerning because they indicate environmental or dependency resolution problems that can prevent the tool from starting at all. These are not feature-request issues; they are blocker-class bugs that prevent users from using the software.
Takeaways
-
Aider is in a high-engagement, low-velocity phase. 61 issues opened in 30 days against 1 commit and 0 merged PRs suggests the project is receiving sustained user attention but the core maintainer(s) are either bottlenecked, triaging rather than shipping, or managing a large backlog. The single-contributor 30-day window is a red flag for bus-factor risk.
-
Import and runtime errors dominate the issue queue. The prevalence of uncaught exceptions in utility modules (_legacy_response.py, _types.py, utils.py) suggests either a recent refactor that introduced regressions or dependency conflicts that vary by installation environment. This is not a feature maturity problem; it is a stability problem.
-
Claude Sonnet 4.5 integration is the active development focus. The single commit enables overeager mode for Anthropic’s latest model, confirming that Aider tracks new LLM releases closely. However, the lack of a formal release to ship this change means users must either track main or wait for a future release cut.
-
No release cadence in 90 days creates friction for production adoption. Projects with 43k stars typically maintain quarterly or monthly releases. The absence of any release in the past 90 days, combined with minimal commit velocity, suggests either a transition period (e.g., new maintainer, major refactor) or a shift away from active feature development. For users evaluating Aider for critical workflows, this is a signal to pin a specific commit and manage upgrades manually.
Further reading
- Aider GitHub repository: The official source code and issue tracker for the AI pair programming tool.
- Aider documentation: Official user guide and API reference for terminal-based AI-assisted development.
- Anthropic Claude API documentation: Reference for Claude Sonnet 4.5 model capabilities and configuration options.
- Python import system documentation: Technical reference for understanding and debugging Python import errors.
- GitHub issue tracking best practices: Guide to managing high-volume issue queues effectively.
Frequently asked
What is Aider and what does it do?
Aider is an open-source terminal-based AI pair programmer that integrates large language models (Claude, GPT-4, Gemini, Llama) into local development workflows. It allows developers to write and refactor code collaboratively with an AI assistant directly in the command line.
Why are there so many import errors in recent issues?
Recent issues show a pattern of uncaught ImportErrors in modules like _legacy_response.py, _types.py, and utils.py. These suggest either a recent refactor that introduced regressions or dependency resolution conflicts that vary by installation environment. Some have been closed quickly, but others remain open.
How often does Aider release new versions?
Based on the latest data, no releases were cut in the past 90 days despite 43k GitHub stars and active issue volume. This suggests users may be running from the main branch or relying on older stable versions, creating a gap between development and production deployments.
What is overeager mode and why was it enabled for Claude Sonnet 4.5?
Overeager mode is a configuration parameter that controls how aggressively the LLM proposes code changes. The recent commit enabled it specifically for Claude Sonnet 4.5 models, suggesting Aider is tuning its behavior to match the capabilities and characteristics of Anthropic's latest model release.
Is Aider actively maintained?
The project shows mixed signals. It has 43k stars and receives high issue volume (61 opened in 30 days), but only 1 commit and 0 merged PRs in the same period, with a single contributor. This suggests either a bottleneck in the maintainer capacity or a transition period, making it important to evaluate bus-factor risk before adopting for critical workflows.