dsh-compact-and-branch
Run the following command in DeepSeek Harness:
dsh plugin install zeropointnine/dsh-compact-and-branch
Paste the following prompt into your AI chat to install this plugin:
In DeepSeek Harness, you can install this plugin from the GitHub repository https://github.com/zeropointnine/dsh-compact-and-branch by executing the command `dsh plugin install zeropointnine/dsh-compact-and-branch`.
About this plugin
When running long coding sessions in DeepSeek Harness, context windows can fill up quickly, necessitating frequent history compaction. While the built-in /compact command is useful, sometimes it's better to treat compaction as a clean branch point: continue from the compacted state in a fresh session while keeping the original session independently resumable.
The dsh-compact-and-branch plugin addresses this by providing the /compact-and-branch command. It compacts the current session's context and creates a brand-new root session with the compacted context, opening it in the browser. The original session remains intact and resumable, allowing for seamless continuation without data loss.
This plugin is ideal for developers using DeepSeek Harness with local LLMs or limited context windows. It streamlines workflow by turning compaction into a branching operation, reducing the need for repeated compactions and making session boundaries less disruptive.
Use Cases
- When context window runs out during long coding sessions
- Need to branch to a new session after compaction to continue work
- Managing session history in local LLM environments
Best For
- Developers using DeepSeek Harness
- Programmers running long coding sessions
- Users of AI agents in local LLM environments
Related Plugins
Traceable, searchable cross-session memory for AI agents that turns conversation knowledge into a typed knowledge graph and recalls relevant subgraphs instead of replaying full history, natively integrated with DeepSeek Harness.
Gives DSH AI cross-session long-term memory, to-do and skill management, plus multi-session orchestration, external AI delegation, and an infinite canvas that grows with you.
dsh-mnemon is a three-tier, pluggable, Agent-driven memory system for DeepSeek Harness, combining Runtime memory, Project Documents, and replaceable Memory Spaces with nine long-term providers.
An opinionated, zero-infrastructure file-based memory protocol: human-readable markdown files plus a bounded always-loaded index, curation discipline, and no database or embeddings.