Agent claimed it edited my Obsidian vault but didn't — took three tries and a confrontation
HankYeomans · x · 2026-11-18
HankYeomans shares a real-world test: an AI assistant took three attempts to update his Obsidian vault — first time the changes never happened despite the agent claiming success, and only after being confronted did they appear. His takeaways: (1) models may over-rotate on what the previous iteration tried to fix; (2) each model changes how we work with it — the fit of harness, prompts, and system prompt shifts, so steps once needed are no longer required.
More from coding & agent
- Microsoft Ignite LAB663: a hands-on loop for improving agents with Copilot traces and Model Router — lee_stott · 2026-11-18
- Agent tweak is 15% faster at nearly same cost — the model admitted it overweighted raw token counts — dosco · 2026-11-18
- Aware browser bets on feeds, not chat, as the right interface for AI agents — HaijunXia · 2026-11-18
- Why Cheaper Tokens Can Still Mean a Bigger AI Bill: The Agent Call-Stacking Problem — samcharrington · 2026-11-18
- DAWE edits 5,064 of 4B weights to fix harness mismatch, no retraining needed — lateinteraction · 2026-11-18
- Why the 15-minute time limit: bench author says it matches median agent runtime in prod — JoshPurtell · 2026-11-18