Every Claude/Codex session produces great work… and then it disappears into scrollback.
I built context-ledger to record & analyze AI work sessions so you can actually reuse your AI context.
Your AI memory, analyzed. Locally.
github.com/manthan787/con…
Seattle
Joined July 2009
- As someone already using AI coding agents (Claude, Codex, etc.), how do you rate the maturity of Ops-focused AI tools (runbook automation, incident response) in comparison?
- I got tired of paying Zapier for automations an LLM could build in seconds. So I built openweavr.ai — an open-source self-hosted AI automation engine. Describe a workflow in plain English → AI generates it → runs on your own hardware. Works with Claude / GPT / Ollama

