Cognition on Agent Engineering Practices
petergyang · x · 2026-07-12
This episode interviews @imjaredz (Builder in Residence at Cognition), discussing how to learn from top engineers to build better AI agents. Key points: - Fewer rules work better: Don't over-constrain the agent; give it more room. - Let agents self-check: Design workflows for models to review their own outputs. - Split tasks: Break work into multiple sub-agents with smaller context windows. - Agents managing agents: Live demo of one agent managing a group of agents. Also, Cognition's goal is not to maximize token usage but ROI; now more Devin sessions are initiated by other Devins than by humans.
Related event: Cognition Shares AI Agent Engineering Insights(3 posts)→
More from coding & agent
- Coding agents are heading toward an AI-writes, AI-reviews, human-approves workflow — aftahi_ai · 2026-07-22
- oMLX 0.5.2 adds Mac menu-bar stats, low-bit decode kernels, and faster downloads — awnihannun · 2026-07-22
- GitHub review bot hits its PR limit and forces a 39-minute cooldown — DanielLockyer · 2026-07-22
- Max reasoning effort appears to be mobile-only in Codex Remote, not desktop — GabGarrett · 2026-07-22
- A Reddit demo argues online stores should expose carts and pricing through MCP — gelembjuk · 2026-07-22
- Open-source AI SDK provider routes Vercel apps through a local Codex subscription — lgrammel · 2026-07-22