Join our Newsletter — 33% off our NHI Course

Context compaction or action ledgers for long-horizon agents?

 

(@nhi-mgmt-group)
Member Moderator
Joined: 2 years ago
Posts: 20225
Topic starter  

TL;DR: Long-horizon agent failures split into four degradation modes, and compaction only helps one of them, according to Agentic Fabriq. Authority state, task state, and completed actions need to stay outside summarised context if agents are to recover reliably.

Editorial analysis by NHI Mgmt Group, based on content published by Agentic Fabriq: “Context Engineering for Long-Horizon Agents”.

Key questions

Q: Why can summarisation make an AI agent less reliable?

A: Because summarisation can strip away provenance and preserve a wrong fact in cleaner form.

Q: How should security teams decide what to keep outside an agent’s summarised context?

A: Keep anything that defines authority or reversibility outside the summarised region.

Q: What should teams do when a long-running AI agent keeps repeating old actions?

A: Treat that as context distraction, not a general failure of the model.

Practitioner guidance

  • Separate durable state from disposable context Keep the task statement, approval state, and action ledger outside any summarisation path so the agent cannot rewrite them during compaction.
  • Classify the failure mode before applying compaction Decide whether the problem is poisoning, distraction, confusion, or clash, then apply the matching control instead of summarising by default.
  • Clear tool outputs as soon as they are harvested Replace large raw tool dumps with the extracted answer and a pointer to the source so the window keeps signal without preserving unnecessary bulk.

Bottom line: Long-horizon agents fail in multiple distinct ways, so a single summarisation strategy cannot be the default response.

Glossary terms: Context Poisoning, Context Compaction, Action ledger, Context distraction

Explore further

View Full Forum →  |  NHI Foundation Course →  |  Our Services →  |  Read the full analysis →


This topic was modified 1 week ago by NHI Mgmt Group

   
Quote
(@mr-nhi)
Member Moderator
Joined: 6 months ago
Posts: 20715
 

Context governance is now a control problem, not a prompt-writing problem. Long-horizon agents fail because the window contains competing claims, stale history, and degraded salience, not because the model simply needs more tokens. That means programme owners need to treat context state as governed evidence, not disposable chat history. The practitioner implication is clear: separate authoritative task state from transient working context.

A question worth separating out:

Q: What is the difference between context poisoning and context clash?

A: Context poisoning is when a wrong fact enters the window and gets reused as if it were true. Context clash is when two incompatible facts remain in the window at the same time. The first problem is about false authority, while the second is about unresolved contradiction, and they need different handling.

👉 Read our full editorial: Long-horizon agent controls must separate authority from memory


This post was modified 1 week ago by NHI Mgmt Group

   
ReplyQuote
Share:

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.