Study: Agents read instructions/notes 60.5% of the time, rarely touch API docs
dair_ai · x · 2026-08-24
A study analyzing 557 agentic coding sessions and 94K development events reveals key behaviors:
- Reading habits: Instruction files and working notes account for 60.5% of what agents read, while classical technical docs make up 10.6% and API references only 1.3%.
- Testing correlation: Reading docs is associated with less immediate testing (adjusted odds ratio of 0.39).
- Consultation triggers: 70.2% of consultations are self-initiated, compared to 7.5% driven by failure.
- Edit patterns: In multi-commit agentic PRs, code is touched first 4.7x more often than other files.
More from coding & agent
- OpenAI Showcases Hands-Free Coding Using Voice in Codex — OpenAIDevs · 2026-08-24
- Swarms: Enterprise-Grade Multi-Agent Orchestration Framework Open Sourced — KyeGomezB · 2026-08-24
- Claude Verifies 43 Lean Modules autonomously, Tackling Theoretical Physics — Tkaraletsos · 2026-08-24
- Compound Engineering Plugin Rewrite Cuts Context Usage by 70% — iamrobotbear · 2026-08-24
- Lighter's prover challenge breaks 100k TPS, 10x above launch baseline in weeks — econoar · 2026-08-24
- AI fakes memory: why it gets confidently wrong without forgetting — PrajwalTomar_ · 2026-08-24