The Tramya Blog
Secrets, memory and control for your AI agents.
Practical, honest writing for developers who work with AI coding agents every day. No hype, no fluff.
Guide · 7 min
Why switching AI coding tools loses your context — and how to keep it
Move from Claude Code to Codex or Cursor and your project history stays behind. Why the context is trapped inside the tool, and how to make memory and secrets portable.
August 24, 2026Guide · 7 min
How prompt injection turns your AI agent into a secrets leak
An agent that reads files and makes network calls can be tricked into exfiltrating your keys. How the attack works — and how to make the secret unavailable to steal.
August 24, 2026Guide · 7 min
Keep context across AI coding sessions without re-explaining everything
Your agent forgets everything between sessions. A persistent, tool-agnostic memory it reads and writes itself — so the next session starts already knowing the last decision.
August 24, 2026Guide · 8 min
How to give AI coding agents API keys without exposing them
Agents read your .env and leak secrets. A safer model: reference the credential, approve each use locally, and inject it into the subprocess, masked, so the model never sees it.
August 24, 2026