Cody Agent

Generates code by querying Sourcegraph's Advanced Search API to retrieve symbol definitions, usage patterns, and API signatures from the entire codebase, then passing this structured context to an LLM. Uses @-syntax to allow users to explicitly pin files, symbols, or remote repositories as context anchors, enabling the LLM to generate code that matches existing patterns and conventions without manual context copying.

Provides real-time code suggestions as users type, using the open file and repository context to generate completions. Implements Context Filters feature that allows teams to exclude specific repositories from autocomplete results, preventing suggestions that reference deprecated or out-of-scope code. Suggestions appear inline in the editor and can be accepted or dismissed without interrupting the user's workflow.

Generates unit tests for code by analyzing the function signature, implementation, and usage patterns in the codebase. Uses Sourcegraph's symbol search to understand dependencies and mocking requirements, then generates tests with appropriate assertions, mocks, and fixtures. Generated tests follow the codebase's existing testing patterns (e.g., test framework, assertion style, fixture organization). Tests are generated as code snippets that users can review and integrate into their test suite.

Analyzes code for potential bugs, logic errors, and security vulnerabilities by examining the code in context of the codebase's patterns and dependencies. Uses Sourcegraph's symbol search to understand how code is used elsewhere and identify potential issues (e.g., null pointer dereferences, SQL injection, race conditions). Generates suggestions for fixes with explanations of the vulnerability and recommended remediation. Vulnerability detection is static analysis only; no runtime execution or dynamic analysis.

Suggests refactorings (e.g., extract function, rename variable, simplify logic) by analyzing code in context of the entire codebase. Uses Sourcegraph's symbol search to understand the impact of proposed changes on dependent code, ensuring that refactorings don't break other parts of the system. Generates refactoring suggestions as diffs that users can review and apply. Refactoring is limited to structural changes; no semantic transformations or algorithm changes.

Implements a data handling policy where prompts and responses from Sourcegraph.com users are NOT used to train or improve Cody's underlying LLM. Data is collected for product improvement and debugging, but is not fed back into model training. Self-hosted and enterprise deployments have full control over data handling. Policy is documented and enforced at the infrastructure level, not just contractually.

Provides a chat interface where users ask questions about code and receive responses grounded in codebase context. Users can pin context using @-syntax to reference specific files, symbols, remote repositories, or non-code artifacts (documentation, design docs). The chat maintains conversation history within a session and retrieves relevant code context automatically based on the query, then passes both conversation history and pinned context to the LLM for response generation.

Monitors cursor movements and typing patterns to detect when a user is editing code, then analyzes the changes in context of the surrounding codebase to suggest fixes, refactorings, or improvements. Uses Sourcegraph's symbol search to understand the impact of changes across the codebase and generates suggestions that account for dependent code. Suggestions are presented as diffs that users can review and apply with a single action.

Allows users to define and execute reusable prompts that perform multi-step code analysis or generation tasks. Provides a library of premade prompts (e.g., 'explain code', 'generate tests', 'find bugs') that can be customized with parameters. Prompts are executed in the context of the current file or pinned context, and results are returned as text, code, or structured data. Execution is synchronous and user-initiated; no background or scheduled execution.

Retrieves code context from multiple repositories using Sourcegraph's Advanced Search API, enabling Cody to understand code patterns, API signatures, and usage examples across the entire codebase. Search is symbol-aware, returning definitions, usages, and related code rather than simple text matches. Results are ranked by relevance and integrated into the LLM context automatically or via explicit @-references. Supports both local and remote repositories.

Allows teams to define Context Filters that exclude specific repositories from autocomplete suggestions and chat context retrieval. Filters are applied globally to all team members and prevent Cody from suggesting code patterns from deprecated, legacy, or out-of-scope repositories. Filtering is repository-level only; no finer-grained filtering by file path, directory, or code ownership is supported.

Provides Cody as extensions for VS Code, JetBrains IDEs (IntelliJ, PyCharm, WebStorm, etc.), and Visual Studio, with native UI components that integrate seamlessly into each editor's interface. Chat, autocomplete, and fix suggestions appear in editor-native panels and inline decorations. Also available as a web UI in Sourcegraph and as a CLI tool. Deployment is consistent across platforms; context retrieval and LLM calls are routed through Sourcegraph backend.

Cody is available for deployment in multiple modes: cloud-hosted on Sourcegraph.com, single-tenant cloud, or self-hosted on customer infrastructure. Enterprise deployments include dedicated support, custom model configuration (if available), and integration with enterprise Sourcegraph instances. Data handling policies differ by deployment mode: cloud users' data is not used for model training, while self-hosted deployments have full data control.

Analyzes code in the context of the surrounding codebase and generates natural language explanations of what the code does, why it's structured that way, and how it integrates with other parts of the system. Uses Sourcegraph's symbol search to understand dependencies and patterns, then generates explanations that reference actual usage examples from the codebase. Can also generate documentation (docstrings, comments, README sections) that matches the codebase's existing documentation style.