10 Free Repos That Teach AI Agent Harness Engineering: Tools, Memory, Rules
blaizedsouza · x · 2026-11-19
A retweet introducing "harness engineering": building the tools, context, memory and rules around an AI agent so it works reliably every time. The author curates 10 free repos as a learning path:
- awesome-harness-engineering (ai-boost): the front door to the field — a curated list covering frameworks, evals, memory, MCP, permissions, observability and orchestration (5.2k stars)
- SWE-agent: a research coding agent whose harness, prompts and tools you can read and modify
- opencode: a coding agent supporting 75+ model providers — swap the model, keep the harness
- awesome agent harness (RUCAIBox): a survey sorting agent harnesses into three layers
- citadel: a harness for Claude Code and Codex with isolated worktrees, multi-agent coordination and persistence
The author wonders why everyone isn't using this yet, and links an article on building a "second brain" with Claude and Obsidian that gets smarter with every source saved.
More from coding & agent
- Kilo adds linked billing with Claude Platform, no API key sharing needed — anacondainc · 2026-11-19
- Google paper: agent harnesses evolved on one model underperform when moved to another — blaizedsouza · 2026-11-19
- Stanford team introduces DiscoveryAgent for rigorous open-ended scientific discovery — james_y_zou · 2026-11-19
- Solo dev cuts sandbox cold-start to 0.45s, sleeping VMs cost 1¢/month — emilahlback · 2026-11-19
- Team cuts agent sandbox cost 88% by moving to GKE, open-sources gke-sandboxes — emilahlback · 2026-11-19
- Google Antigravity shipped 150 releases in 90 days — a Googler's hands-on review finds it finally caught up — rseroter · 2026-11-19