Cody, an AI coding assistant by Sourcegraph, is designed to enhance developer productivity through advanced search and codebase context integration within IDEs like Visual Studio Code. It offers features such as semantic search, autocomplete suggestions, and unit test generation.
The project has seen significant activity with 180 open issues, many of which highlight user frustrations regarding the limitations of the Pro plan, particularly concerning rate limits and functionality discrepancies. Recent issues also focus on bugs affecting user experience, such as context retrieval problems and UI stability issues. The development team is actively addressing these concerns with numerous commits aimed at improving both frontend and backend functionalities.
Recent issues and pull requests indicate a focus on resolving bugs and enhancing user experience. Key issues include syntax highlighting errors (#5300), discrepancies in commit message features (#5298), and UI problems like the chat window displaying only a gray background (#5296). These issues suggest a need for improved stability and clarity in feature offerings.
The Cody project is actively evolving, with a strong emphasis on addressing user feedback to enhance usability and reliability across various coding tasks within integrated development environments.
| Timespan | Opened | Closed | Comments | Labeled | Milestones |
|---|---|---|---|---|---|
| 7 Days | 13 | 24 | 4 | 0 | 1 |
| 30 Days | 49 | 88 | 62 | 1 | 1 |
| 90 Days | 159 | 170 | 283 | 7 | 1 |
| All Time | 1651 | 1471 | - | - | - |
Like all software activity quantification, these numbers are imperfect but sometimes useful. Comments, Labels, and Milestones refer to those issues opened in the timespan in question.
| Developer | Avatar | Branches | PRs | Commits | Files | Changes |
|---|---|---|---|---|---|---|
| Rik |  |
6 | 10/4/1 | 59 | 93 | 29437 |
| Quinn Slack |  |
8 | 40/35/1 | 98 | 379 | 27708 |
| Beatrix |  |
12 | 27/24/1 | 45 | 265 | 8975 |
| Tom Ross |  |
4 | 22/19/1 | 33 | 90 | 8826 |
| Beyang Liu |  |
6 | 14/12/1 | 29 | 55 | 4797 |
| Ólafur Páll Geirsson |  |
3 | 10/8/1 | 16 | 60 | 4159 |
| Valery Bugakov |  |
2 | 6/6/0 | 10 | 65 | 2257 |
| Vova Kulikov |  |
2 | 12/12/0 | 16 | 60 | 2067 |
| James McNamara |  |
7 | 5/4/0 | 19 | 58 | 2000 |
| Piotr Kukiełka |  |
1 | 4/4/0 | 6 | 55 | 1858 |
| Naman Kumar |  |
2 | 5/4/0 | 11 | 51 | 1396 |
| Dominic Cooney |  |
2 | 0/0/0 | 8 | 31 | 1363 |
| Ara (arafatkatze) | 4 | 3/0/2 | 14 | 19 | 905 | |
| Keegan Carruthers-Smith |  |
2 | 2/2/0 | 9 | 34 | 598 |
| None (renovate[bot]) | 1 | 1/0/0 | 1 | 1 | 566 | |
| Tim Lucas |  |
2 | 0/0/0 | 2 | 20 | 488 |
| Hitesh Sagtani |  |
2 | 10/8/1 | 7 | 8 | 390 |
| Mikołaj Kondratek |  |
4 | 7/3/3 | 7 | 14 | 367 |
| Aditya Kalia |  |
1 | 4/4/0 | 5 | 14 | 364 |
| Justin Milner |  |
2 | 1/0/1 | 2 | 11 | 330 |
| Jean-Hadrien Chabran |  |
2 | 3/2/0 | 12 | 6 | 278 |
| Daniel Marques |  |
1 | 0/0/0 | 7 | 5 | 263 |
| Kevin Chen |  |
5 | 4/1/1 | 11 | 9 | 217 |
| Camden Cheek |  |
1 | 1/1/0 | 1 | 7 | 182 |
| Jan Hartman |  |
2 | 2/1/0 | 8 | 11 | 153 |
| Kalan |  |
2 | 2/2/0 | 3 | 2 | 30 |
| Rafał Gajdulewicz |  |
1 | 1/1/0 | 1 | 3 | 20 |
| Taras Yemets (taras-yemets) | 1 | 1/0/0 | 1 | 1 | 9 | |
| YK |  |
1 | 1/1/0 | 1 | 1 | 2 |
| M.F.M Fazrin (nirzaf) | 0 | 2/0/1 | 0 | 0 | 0 |
PRs: created by that dev and opened/merged/closed-unmerged during the period
The Cody project on GitHub has seen a significant uptick in activity, with 180 open issues currently logged. Recent issues predominantly report bugs related to functionality and user experience, particularly concerning the integration of AI models and the handling of context in coding tasks. A noticeable trend is the frustration expressed by users regarding the limitations of the Pro plan, particularly concerning rate limits and functionality that seems to contradict advertised capabilities.
Several issues highlight recurring themes: problems with context retrieval, difficulties in model switching, and challenges with the user interface that disrupt workflow. The community appears engaged yet frustrated, signaling a critical need for improvements in both performance and clarity of features.
Issue #5300: bug: Julia has not syntax highlighted
Issue #5298: bug: Experimental Commit Messages Feature Doesn't Match Actual Code Changes
Issue #5296: bug: The Cody chat window panel shows nothing but a gray background.
Issue #5295: bug: Not loading historical chat
Issue #5274: bug: Cody not able to view or access files/context that I attach in the chat.
Issue #5227: bug: commands don't work anymore (Edited 3 days ago)
Issue #5207: bug: Cody panel prevents maximizing terminal panel (Edited 7 days ago)
Issue #5206: bug: 应用重新产生的注释内容，会移除代码，仅展示注释信息 (Edited 7 days ago)
Issue #5204: bug: Unable to login using GitHub (Edited 7 days ago)
Issue #5200: feedback: Ability to see history for edit code prompts (Edited 1 day ago)
A significant number of recent issues relate to bugs in the user interface and functionality, particularly around how context is managed when interacting with code files. Users frequently report that Cody fails to recognize or utilize provided context effectively, leading to erroneous suggestions or incomplete responses.
The issue regarding commit messages (#5298) indicates a potential disconnect between expected and actual functionality, which could undermine user trust in the tool's reliability for version control tasks.
The reoccurring problem of the chat window displaying only a gray background (#5296) suggests possible stability issues within the extension that could deter users from engaging with its features.
User feedback indicates frustration with rate limits on the Pro plan (#4199), highlighting a need for clearer communication about usage limits and expectations for paid users.
The lack of proper syntax highlighting for languages like Julia (#5300) reflects a broader concern about language support within the tool, which may limit its usability for developers working in diverse coding environments.
In summary, while there is active engagement from users reporting issues and providing feedback, there are clear pain points that need addressing to enhance usability and reliability across various coding tasks within Cody.
The Cody project has a total of 49 open pull requests (PRs) with a diverse range of changes aimed at improving functionality, fixing bugs, and enhancing user experience. The recent activity reflects ongoing development efforts, with several PRs addressing specific issues and implementing new features.
PR #5299: Chat: Improve Cody Web initialization and view handling
This PR optimizes the initialization process for the Cody Web interface by reducing duplicated messages and ensuring proper view handling based on authentication status. It addresses a regression that caused the login view to display incorrectly.
PR #5292: PLG: simplify model selection
Simplifies the model selection process by removing the "Balanced" group and refining the available models for users. This change aims to streamline user choices and enhance clarity.
PR #5290: Smart Apply: Use system prompt to encourage code blocks being created
Enhances the smart apply feature by utilizing the system prompt to guide users in creating code blocks, improving overall usability.
PR #5289: Edit: Support multiple models on enterprise
Introduces support for selecting multiple models in enterprise settings, expanding customization options for users.
PR #5287: Fix scrollbar for LLM Dropdown in Cody plugin
Adds a scrollbar to the LLM dropdown in the VS Code plugin, improving accessibility when many options are available.
PR #5281: Codegen: add C# support for Agent bindings
Adds support for generating C# bindings in the code generation process, facilitating integration with Visual Studio.
PR #5278: E2Ev2: Baseline Framework
Introduces a new end-to-end testing framework aimed at improving test reliability and speed, addressing issues with flaky tests in previous iterations.
PR #5276: added(telemetry): separate field for IDE name and server identification
Updates telemetry to better identify IDEs and server interactions, enhancing data collection accuracy.
PR #5272: DO NOT MERGE: demo deepseek setup
A draft PR intended for demonstration purposes related to deepseek setup.
PR #5264: reverting back to accuracy instead of power to not break sg/sg backend
Reverts changes made in a previous PR to restore functionality related to model accuracy.
Several older PRs focus on refining existing features, optimizing performance, and cleaning up code. Notably:
The current set of open pull requests reflects several key themes:
User Experience Improvements: Many recent PRs focus on enhancing user interaction with the Cody interface. For instance, PRs like #5299 and #5287 address UI responsiveness and clarity, which are crucial for maintaining user engagement. The emphasis on fixing regressions (e.g., improper view handling) indicates a commitment to providing a stable user experience.
Feature Expansion: There is a clear trend towards expanding features, particularly in terms of model selection (#5292) and multi-model support (#5289). This aligns with the project's goal of offering diverse AI capabilities tailored to different user needs.
Testing and Reliability Enhancements: The introduction of new testing frameworks (#5278) alongside improvements in telemetry (#5276) shows a proactive approach towards maintaining software quality. The focus on reducing flakiness in tests indicates an understanding of the importance of reliable software delivery.
Community Engagement: The high number of open issues and pull requests suggests an active community contributing feedback and requesting features. This is further supported by discussions around PRs where contributors engage in meaningful dialogue about implementation details and potential improvements.
Technical Debt Management: Several PRs aim to clean up code or remove unused features (e.g., PRs #5246, #5268). This is essential for maintaining code quality as the project scales, ensuring that it remains manageable and efficient over time.
Diversity of Contributions: The variety of contributors working on different aspects—from UI enhancements to backend improvements—demonstrates a collaborative environment that encourages contributions across various skill sets.
In conclusion, the Cody project is actively evolving with a strong focus on user experience, feature expansion, testing reliability, community engagement, technical debt management, and diverse contributions. These factors collectively contribute to its potential success as an AI coding assistant within integrated development environments.
Vova Kulikov (vovakulikov)
Mikołaj Kondratek (mkondratek)
Naman Kumar (thenamankumar)
Hitesh Sagtani (hitesh-1997)
Valery Bugakov (valerybugakov)
Aditya Kalia (akalia25)
Quinn Slack (sqs)
Tom Ross (umpox)
Kevin Chen (chenkc805)
Beyang Liu (beyang)
Summary
Quantified Reports
Detailed Reports