Skip to content

Copilot's Quiet Comeback: GitHub's AI Tool Isn't Done Yet

thoughts

I took the time to summarize GitHub Copilot’s upcoming capabilities — specifically prompt management, Agent mode, and MCP tools extension support. The question: is this enough for GitHub to maintain its competitive advantage in the enterprise market?

What’s changed

Prompt Management — Three-tier system: global custom instructions in .github/copilot-instructions.md, action-specific instructions for different tasks (code generation, testing, commits), and reusable prompt snippets stored in .github/prompts.

Agent Mode — Autonomous reasoning capabilities with continuous conversation, code planning, and model flexibility including Claude 3.5 alternatives. This was the piece that was previously missing.

MCP Tool Integration — Docker, npm, and PyPI servers with automatic tool discovery. There’s even a “Yolo mode” (chat.tools.autoApprove: true) for automatic execution.

The real question

While acknowledging commoditization of features across competing tools, four improvement directions matter:

  1. Transitioning from prompt management to specification writing
  2. Improving decision-making UX
  3. Supporting parallel exploration
  4. Converting code into learning opportunities

The enterprise angle is real — procurement barriers slow adoption of alternatives like Cursor and Windsurf, giving Copilot time to close the gap.


Full article at tessl.io. Originally posted on LinkedIn.