Supported AI Coding Agents

This document provides a comprehensive list of AI coding agents and assistants that the Coding Context CLI can integrate with. The tool is designed to be agent-agnostic, but provides specific support for popular agents’ configuration file locations and conventions.

Currently Supported Agents

The following agents have explicit support with dedicated configuration file paths:

1. GitHub Copilot

  • Website: https://github.com/features/copilot
  • Provider: GitHub (Microsoft)
  • Config Locations:
    • .github/copilot-instructions.md
    • .github/agents
  • User Rules Path: ~/.github/agents/AGENTS.md
  • Description: AI pair programmer integrated into VS Code, Visual Studio, JetBrains IDEs, and Neovim
  • Agent Flag: -a copilot

2. Anthropic Claude

  • Website: https://claude.ai/
  • Provider: Anthropic
  • Config Locations:
    • CLAUDE.md
    • CLAUDE.local.md
    • .claude/CLAUDE.md
  • User Rules Path: ~/.claude/CLAUDE.md
  • Description: Large language model with extended context windows, accessible via API and web interface
  • Agent Flag: -a claude

3. Cursor

  • Website: https://cursor.sh/
  • Provider: Cursor
  • Config Locations:
    • .cursor/rules
    • .cursorrules
    • .cursor/commands
  • User Rules Path: ~/.cursor/rules/AGENTS.md
  • Description: AI-first code editor built on VS Code with integrated AI chat and editing
  • Agent Flag: -a cursor

4. Google Gemini

  • Website: https://gemini.google.com/
  • Provider: Google
  • Config Locations:
    • GEMINI.md
    • .gemini/styleguide.md
  • User Rules Path: ~/.gemini/GEMINI.md
  • Description: Google’s multimodal AI model with coding capabilities
  • Agent Flag: -a gemini

5. Augment

  • Website: https://augmentcode.com/
  • Provider: Augment
  • Config Locations:
    • .augment/rules
    • .augment/guidelines.md
  • User Rules Path: ~/.augment/AGENTS.md
  • Description: AI coding assistant with contextual understanding
  • Agent Flag: -a augment

6. Windsurf

  • Website: https://codeium.com/windsurf
  • Provider: Codeium
  • Config Locations:
    • .windsurf/rules
    • .windsurfrules
  • User Rules Path: ~/.windsurf/AGENTS.md
  • Description: AI-powered IDE from Codeium
  • Agent Flag: -a windsurf

7. OpenCode.ai

  • Website: https://opencode.ai/
  • Provider: OpenCode.ai
  • Config Locations:
    • .opencode/agent
    • .opencode/command
    • .opencode/rules
  • User Rules Path: ~/.opencode/AGENTS.md
  • Description: Open-source AI coding platform
  • Agent Flag: -a opencode

8. Codex

  • Website: https://codex.ai/
  • Provider: Codex
  • Config Locations:
    • AGENTS.md
    • .codex/AGENTS.md
  • User Rules Path: ~/.codex/AGENTS.md
  • Description: AI coding assistant platform
  • Agent Flag: -a codex

Additional Agents to Consider

These agents are widely used but may not yet have explicit configuration path support:

AI-Powered IDEs and Editors

9. Codeium

  • Website: https://codeium.com/
  • Provider: Codeium
  • Description: Free AI code completion tool with support for 70+ languages
  • Supported IDEs: VS Code, JetBrains, Vim/Neovim, Emacs, Eclipse, and more
  • Note: Parent company of Windsurf

10. Tabnine

  • Website: https://www.tabnine.com/
  • Provider: Tabnine
  • Description: AI code completion that can run locally or in the cloud
  • Supported IDEs: VS Code, JetBrains, Vim, Sublime, Atom, and more
  • Privacy: Offers local model option for sensitive code

11. Amazon CodeWhisperer (now Amazon Q Developer)

  • Website: https://aws.amazon.com/q/developer/
  • Provider: Amazon Web Services
  • Description: AI coding companion with AWS service integration
  • Supported IDEs: VS Code, JetBrains, AWS Cloud9, AWS Lambda console

12. Replit Ghostwriter

  • Website: https://replit.com/
  • Provider: Replit
  • Description: AI pair programmer integrated into Replit’s online IDE
  • Features: Code completion, generation, and debugging

13. Sourcegraph Cody

  • Website: https://sourcegraph.com/cody
  • Provider: Sourcegraph
  • Description: AI coding assistant with codebase context awareness
  • Supported IDEs: VS Code, JetBrains, Neovim
  • Features: Code search, understanding, and generation

Standalone AI Coding Tools

14. OpenAI GPT-4 / ChatGPT

  • Website: https://openai.com/
  • Provider: OpenAI
  • Description: General-purpose LLM with strong coding capabilities
  • Access: Web interface, API, and third-party integrations

15. GPT-Engineer

16. Aider

  • Website: https://aider.chat/
  • Provider: Open Source
  • Description: AI pair programming in your terminal with Git integration
  • Features: Direct file editing, Git commits, works with various LLMs

17. Continue

  • Website: https://continue.dev/
  • Provider: Open Source
  • Description: Open-source autopilot for VS Code and JetBrains
  • Features: Chat, edit, and generate with any LLM

18. Copilot++ / Copilot Workspace

  • Website: https://githubnext.com/
  • Provider: GitHub Next (Microsoft)
  • Description: Experimental features and future of GitHub Copilot

Cloud-Based Development Environments

19. GitHub Codespaces

  • Website: https://github.com/features/codespaces
  • Provider: GitHub (Microsoft)
  • Description: Cloud-based development environment with Copilot integration
  • Features: Pre-configured containers, VS Code in browser

20. GitLab Duo

  • Website: https://about.gitlab.com/gitlab-duo/
  • Provider: GitLab
  • Description: AI-powered features across GitLab platform
  • Features: Code suggestions, chat, vulnerability explanations

Agent Frameworks and Platforms

21. LangChain

  • Website: https://www.langchain.com/
  • Provider: LangChain
  • Description: Framework for developing LLM-powered applications
  • Use Case: Build custom coding agents

22. AutoGPT

23. BabyAGI

Specialized Coding Assistants

24. Phind

  • Website: https://www.phind.com/
  • Provider: Phind
  • Description: AI search engine for developers with coding focus

25. Bard (now Gemini)

  • Website: https://bard.google.com/
  • Provider: Google
  • Description: Conversational AI with code generation (now merged into Gemini)

26. Perplexity AI

27. Blackbox AI

28. CodeGPT

  • Website: https://codegpt.co/
  • Provider: CodeGPT
  • Description: AI assistant for developers in IDE

29. Pieces for Developers

  • Website: https://pieces.app/
  • Provider: Pieces
  • Description: AI-powered code snippet manager and workflow tool

30. Mintlify

Enterprise and Specialized Solutions

31. Codegen (by Salesforce)

  • Provider: Salesforce Research
  • Description: Open-source code generation models

32. StarCoder / BigCode

33. WizardCoder

  • Provider: WizardLM team
  • Description: Code-focused LLM fine-tuned from CodeLlama

34. DeepSeek Coder

35. Mistral Codestral

  • Website: https://mistral.ai/
  • Provider: Mistral AI
  • Description: Code-specialized model from Mistral

Agent Support Tiers

Tier 1: Full Support

Agents with dedicated configuration paths and -a flag support:

  • GitHub Copilot
  • Claude
  • Cursor
  • Gemini
  • Augment
  • Windsurf
  • OpenCode.ai
  • Codex

Tier 2: Compatible

Agents that can use the tool via standard input/output but lack specific configuration support:

  • Codeium
  • Tabnine
  • Amazon Q Developer
  • Sourcegraph Cody
  • Continue
  • Aider
  • All LLM APIs (OpenAI, Anthropic, Google, etc.)

Tier 3: Framework Integration

Agent frameworks that can incorporate this tool:

  • LangChain
  • AutoGPT
  • BabyAGI
  • Custom agent implementations

Adding Support for New Agents

To add full support for a new agent:

  1. Define configuration paths in the codebase (pkg/codingcontext/paths.go)
  2. Add agent constant to the agent types
  3. Define user rules path for the -w flag
  4. Update documentation in README.md and this file
  5. Add tests for the new agent’s configuration discovery

Example configuration:

// In agent.go
const (
    AgentNewAgent Agent = "newagent"
)

// In paths.go
var newAgentPaths = []string{
    ".newagent/rules",
    ".newagentrules",
}

Recommendations

For Individual Developers

  • GitHub Copilot or Cursor: Best IDE integration
  • Claude or ChatGPT: Best for complex problem-solving
  • Aider: Best for terminal-based workflows

For Teams

  • GitHub Copilot: Best for Microsoft ecosystem
  • Cursor: Best for AI-first coding experience
  • Sourcegraph Cody: Best for large codebases
  • Amazon Q Developer: Best for AWS-heavy projects

For Privacy-Conscious Users

  • Tabnine (local mode): Run models locally
  • Continue: Open-source, use any LLM
  • Aider: Open-source, Git-integrated

For Research and Experimentation

  • OpenAI GPT-4: Most capable general model
  • Claude: Largest context window
  • Open-source models: StarCoder, DeepSeek Coder, WizardCoder

See Also