Context engineering · Agent handoff

Autocompaction is not memory

Built-in context compaction helps an individual agent survive a long chat. A local handoff control plane helps a team preserve the operational state that the next agent needs to act safely.

Published May 26, 2026 · Gregory Shevchenko · Humanswith.ai

Short answer: larger context windows and automatic chat summaries delay context failure. They do not make state portable, auditable, or shared across Claude Code, Codex, Cursor, Windsurf, a remote Mini, and background MCP workflows.

What built-in autocompaction does well

Long-running coding agents already summarize. That is useful. A product-level compaction pass can keep a conversation moving after the raw context becomes too large.

But the summary is usually local to one product session. It is optimized for keeping the current conversation alive, not for handing the work to a different agent surface with approvals, risks, and exact values intact.

The missing layer is operational handoff

A team workflow needs something stricter than a narrative summary after context pressure is already high. It needs a handoff contract that preserves what the next agent must know before it acts.

For us, that means a local handoff MCP that can preserve:

the objective and done condition;
loaded instructions and constraints;
approval state;
exact values that must not drift;
risks and red flags;
actions already taken;
errors and fixes;
the next recommended step;
what the next agent should not redo.

Autocompaction vs local handoff MCP

Capability	Built-in autocompaction	Local handoff MCP
Scope	Usually one product session	Shared workspace protocol
Timing	Often after context pressure is high	Pre-score before the window is full
State	Narrative summary	Operational contract with approvals and risks
Portability	Product-specific	Claude Code, Codex, Cursor, and Windsurf vocabulary
Measurement	Usually opaque	Resume success, re-read rate, token estimate, and leak rate

Why 1M context is not enough

A 1M context window is valuable. It lets the agent keep more code, logs, source material, and prior reasoning available before it needs compression.

But it does not solve the coordination problem. A larger window delays the failure mode. It does not automatically mark which facts are trusted, which approvals were granted, which values must not change, or what a fresh agent should avoid redoing.

The real product is not compression. It is continuity.

The mechanism we are building

Pre-score before the context window is full.
Trigger red flags when approvals, secrets, live ops, or exact values are at risk.
Write a structured handoff contract, not a vibe summary.
Resume with a blind gate: can a fresh agent continue from the handoff alone?
Measure resume success, re-read rate, token estimate, and leak rate.