Observed arrival · 2026-10-16
OpenAgentd lets you watch a coding agent work
An open-source desktop coding agent that runs on your machine and displays its tool calls, code changes, and reported cost.
- For
- Developers who want to inspect agent-driven code changes
- Worth noticing
- Its example replay shows a test failing before the fix, then reports 10 passing tests after the change.
Field notes
The homepage's replay walks through a Python inventory bug: it searches for check_stock, reads inventory and test files, adds a test that fails, changes the implementation, then reports 10 passing tests. The feature list also describes steering during a turn, background helper agents, dev-server browsing, scheduled prompts, plugin hooks, denied-path glob patterns, and MCP/skills connections; these are presented as product capabilities, not independently verified here.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue