Making Docs and AI Work Together: Agentic Coding in the IDE
summary
Built a pipeline that turns API docs into IDE-executable context: Cloudinary MCP servers plus an LLM client in Cursor, with Context7 indexing and short rules files that the model reads before writing. Measured outcome: developers report faster work but only 3.8% show low hallucination with high confidence, so rules and docs-as-context are required. Demo uploads an image, auto-creates a named transformation, applies a text overlay and a g_face+c_auto thumbnail — all generated code runs and the page renders as expected.
AI-written summary · about
chapters (14)
- 0:00Introduction: Setting the Scene
- 1:02Meet the Speaker: Sharon Nik
- 1:22Transformative Experience with Cloudinary APIs
- 2:31Using AI Effectively in the IDE
- 2:58Code Generation and Refactoring with LLM Client
- 4:14Measurable Benefits of AI in Development
- 5:34Challenges and Opportunities in AI Adoption
- 7:35Improving AI Context with Documentation
- 8:05Best Practices for AI Accuracy
- 8:32Demonstration: Adding Cloudinary MCP Servers
- 12:19Creating and Using Rules Files
- 18:13Live Use Case: Cloudinary in Action
- 25:00Future of Documentation and AI
- 25:32Conclusion and Thank You
related talks
23mAI-native CMS vs vibe coding
35mAgentic coding
49mNext-Gen Documentation with AI
01h 00m10 tips to level up your ai-assisted coding
24mHow Agentic AI Empowers Creative Thinking
20mAI editors – how they work, and how to master them
01h 02m10 Tips To Level Up Your AI-Assisted Coding
46mPanel: AI in the Loop - Left-Shifting Platform DX and Governance