· multi-agent-system
GPT-5.6 subagents: context isolation in the weights
GPT-5.6 Sol's Ultra mode gives each subagent its own context window. It solves context isolation but not handoff, and it multiplies your token bill fast.
Blog
The Wire Blog
Practical context engineering for AI agents: how to structure, deliver, and manage the context that determines whether AI systems actually work.
Research-backed articles on context architecture, retrieval, agent security, and making AI tools more effective. Written by Jitpal Kocher.
· Agent Reliability
AI agent reliability is a context problem
AI agent reliability fails because the same task assembles different context every run. Non-determinism is a context engineering problem, not a model flaw.
· Context Pruning
Context pruning helps agents, until it doesn't
Context pruning helps AI agents in one regime and hurts in another. A 2026 study of models from 4B to 284B maps when to prune stale context and when not to.
· RAG (Retrieval-Augmented Generation)
Knowledge graphs vs RAG: when graphs actually win
Knowledge graphs vs RAG: graph retrieval wins multi-hop reasoning and global summarization, but vector RAG matches it on simple lookups at far lower cost.
· Context Bloat
Context bloat: why long-running agents break
Context bloat is when accumulated tool-call output crowds out an agent's task. Tool calls, not window size, break long-running agents. Here is the fix.
· Chunking
Chunking Strategies Decide What Context AI Sees
Chunking strategies decide which context reaches your AI. How fixed-size, semantic, and late chunking change retrieval accuracy, with 2026 benchmark data.
· Context Engineering
How agents manage their own context window
Agent context management is shifting from fixed harness rules to learned, runtime decisions an agent makes about its own window. What the 2026 research shows.
· Context Window
Demand paging for the AI context window
A 2026 systems paper found 21.8% of tokens in agent context windows are wasted. Demand paging treats the AI context window as L1 cache, not full memory.
· MCP (Model Context Protocol)
MCP authorization decides what context agents see
MCP authorization became a context control plane in 2026. RFC 8707 token scoping decides which sources an agent can ever pull into its own context window.
· MCP (Model Context Protocol)
Stateless MCP: the spec is dropping session state
The 2026 MCP release candidate goes stateless: no initialize handshake, no session ID, any server instance answers any request. What stateless MCP means.