AI agents · How Founders Use AI
How do I give my agent the right context and memory so it stops forgetting?
Agents fail less because the model is dumb and more because its context window is stuffed with the wrong things: performance measurably degrades as context grows (context rot). Context engineering is the fix: write important facts to external notes, select only what the current step needs, compress long histories into summaries, and isolate big sub-tasks into fresh contexts. Manus's production lesson adds a money angle: structure context so it caches well, because cache hits are 10x cheaper than fresh tokens.
On the AI map, 29 October 2026 Decision modelsCoding agentsAgents that do tasks for youAI browsersNo-code automation with AI steps
Go deeper18 sources
-
Effective context engineering for AI agents
Anthropic · Guide · 2025
The canonical guide: curation beats clever prompting for agents.
-
Context Engineering for AI Agents: Lessons from Building Manus
Yichao 'Peak' Ji (Manus) · Article · Jul 2025
Battle scars from a real production agent, including the KV-cache economics.
-
Context Engineering for Agents
LangChain · Article · Jul 2025
The write, select, compress, isolate framework every team now uses.
-
Context Engineering for Agents (original essay)
Lance Martin · Article · Jun 2025
The post that coined the framing, with the LLM-as-CPU analogy.
-
Context Engineering in Manus
Lance Martin · Article · Oct 2025
An expert's teardown of how Manus applies the theory in production.
-
Context Engineering for Agents | Lance Martin, LangChain
Latent Space (Lance Martin) · Podcast · Jul 2025
The full conversation, including memory across sessions.
-
Context Engineering for Agents (audio)
Latent Space · Podcast · Jul 2025
Same episode, playable on your commute.
Listen here
All 18 sources
-
Context Engineering for AI Agents: Part 2
Philipp Schmid (Google DeepMind) · Article · 2025
Advanced patterns once the basics stop being enough.
-
Context Engineering Long-Horizon Agents | Harrison Chase
Sequoia Training Data (Harrison Chase) · Podcast · Jan 2026
Why context, not model size, unlocks agents that work for hours.
-
Context engineering in agents (docs)
LangChain · Guide · 2025
The implementation reference when you actually build it.
-
Fighting Context Rot
Inkeep · Article · 2025
Explains why agents get dumber as conversations get longer, and what to do.
-
Diagnosing and Mitigating Context Rot in Long-horizon Search
arXiv researchers · Article · 2026
The research evidence that performance degrades with input length.
-
Claude's Context Engineering Secrets: Best Practices from Anthropic
Bojie Li · Article · Dec 2025
Reverse-engineers how Claude's own products manage context.
-
Context Engineering: A Practical Guide for AI Agents
Sourcegraph · Guide · 2026
How a code-intelligence company feeds agents exactly the right slice of a codebase.
-
AI Agent Harness, 3 Principles for Context Engineering, and the Bitter Lesson Revisited
Hugo Bowne-Anderson · Article · 2025
Connects context engineering to harness design in three principles.
-
Inside the LangChain x Manus Webinar on Context Engineering
Duarte Caldas Cardoso · Article · 2025
Notes from the two leading teams comparing approaches live.
-
Harrison Chase: 2026 ushers in long-horizon agents
36Kr (Harrison Chase interview) · Article · 2026
Where memory and context management are heading next.
-
Context Engineering for AI Agents: Key Lessons from Manus
Context.Space · Article · 2025
A quick digest of the Manus lessons if you skip the original.