Hi, I'm Akshit 👋

Staff Software Engineer, AI Platform @ Cendyn

Building production-grade AI infrastructure: LLM gateways, MCP infrastructure, and agent harnesses. Nine years from frontend architecture to tech lead to Staff. I write about what breaks..

Featured posts

See all posts

Anatomy of an Agent Harness: The Code That Runs When You Type a Prompt

23/08/2026

The LLM never touches your filesystem. A look inside the agent harness, the run loop that turns model output into actual actions, and why understanding it makes you better at working with agents.

Long-Running Agents: Checkpoints, Resume, and Steering

23/08/2026

Ten-hour agent runs die at hour nine for boring reasons. How harnesses use checkpoints, session persistence, and mid-run steering so long autonomous work survives contact with reality.

Diagrams Your AI Agent Can Actually Read

01/07/2026

A ranked guide to diagram formats for repos that AI agents now read alongside humans: ASCII, Mermaid, C4-as-code, and why draw.io should never be your source of truth.

Skills vs. MCP: Understanding the Tools Powering the Next Generation of AI Agents

25/01/2026

Skills provide the procedural 'know-how', while MCP provides the 'tools' to interact with the world. Learn how they differ and how to use them together.

What I'm building

AI Platform & Gateways

Enterprise LLM and MCP gateways. Provider routing, BYOK, Bedrock agent runtimes with typed streaming, Postgres tests with Testcontainers.

FastAPITypeScriptAWS BedrockPostgres

pi-agent-proxy

Native observer for pi. Token audits, tool inventory, and wire transcripts with no HTTP proxy.

TypeScriptPi ExtensionObservability

external-model Agent Skill

Delegate subtasks to cheaper, faster models like DeepSeek, Qwen, and Kimi from Claude Code and other CLIs.

TypeScriptAgent SkillsClaude Code

WebMCP & Agent-Ready Web

Tool manifests served from this site. Machine readable APIs for agents, over web standards.

WebMCPBrowser ToolsStandards