dsh-compaction-policy

dsh plugin --profile web add dsh-compaction-policy

dsh-compaction-policy

English | [中文](README-zh.md)

**Stop DeepSeek Harness from condensing your conversation too early — and put the

trigger where you want it.**

DSH automatically condenses older history when a conversation grows. On a lot of

models it starts much earlier than "80% full", and then it does it again a few

steps later. This plugin gives you that decision back: it does not edit any

config file, it adjusts the running compaction engine.

Is this for you?

Probably yes if any of these sound familiar:

• Your model's context window is smaller than roughly 330,000 tokens (most

local models, many hosted models).

• DSH condenses the conversation while the context ring still looks half empty.
• Right after a condensation the ring climbs back to the same place within a few

steps, so it happens over and over.

If you only ever use very large windows (e.g. 1M), the built-in behaviour is

already near 80% and you most likely don't need this.

Why it happens

Before each step DSH checks how large the next request would be and compares it

to this trigger:

trigger = min(window × 0.8, window − output reserve − 65536)

• window — the model's context window as DSH knows it.
• output reserve — how many tokens the request asks the provider to keep free

for the answer.

• 65536 — a fixed safety margin DSH always sets aside.

The second half of that formula is the catch: it subtracts a fixed 65,536

tokens regardless of how large the window is. Once the window is below about

330,000, that fixed subtraction is the tighter of the two limits, so it becomes

the real trigger:

Context window · Where DSH actually triggers · Where 80% would be

128k · ~62k (49%) · 102k

150k · ~84k (56%) · 120k

256k · ~190k (74%) · 205k

1M · ~678k (68%) · 800k

On very small windows (roughly up to 65k) the fixed margin swallows the whole

window: DSH then refuses to configure proactive compaction at all and only reacts

after a request is rejected for being too long — which throws away far more

history than a planned condensation would.

Why editing a config file doesn't help

In DSH 0.2 the compaction policy lives inside each agent preset, and a preset is

addressed as one whole definition: neither a profile config file nor a plugin's

own config layer can change a single field of it. The only way to write it to disk

is to replace the preset body — which pins your preset and stops it from picking

up DSH updates.

So this plugin doesn't touch any file. It changes the behaviour of the compaction

engine while DSH runs.

What it does

• Below your trigger it answers "nothing to do", so DSH's own check never runs.
• At or above it, the decision goes straight back to DSH — identical to the stock

behaviour.

• Emergency recovery after an over-long request and the manual /compact command

are never touched.

• If anything inside the plugin goes wrong, it steps aside and DSH behaves exactly

as if the plugin weren't installed.

Which DSH versions this works with

DSH version · Status

0.2.2-rc.1 · Verified — the build this release was developed and tested against

later 0.2.x / 0.3.x · Expected to work: the peers are open-ended, and if an internal seam changes the plugin steps aside instead of breaking your session

earlier than 0.2.0-rc.2 · Not supported — it wraps the per-agent-preset compaction engine, which only exists in the 0.2 line

The machine-readable form of this is in package.json: open-ended @deepseek-ai/* peer ranges, with the tested build recorded in dsh.verifiedWith. If it does nothing on your build, nothing breaks — check that Settings lists Compaction policy, then open an issue with dsh --version.

_Last verified: DSH 0.2.2-rc.1, 2026-11-06 (previously 0.2.1-rc.3)._

Install

dsh plugin --profile <your profile> add github:zyd232/dsh-compaction-policy

Use the profile you actually run — desktop for the desktop app, web for

dsh web, or your own profile name. Then restart DSH: plugin rows are only

read at startup, so without a restart nothing changes.

You do not need to start a new conversation. Existing conversations are picked

up on their next step.

To remove it again:

dsh plugin --profile <your profile> remove dsh-compaction-policy

Where the settings are

Open Settings and look at the left sidebar: you will find a **Compaction

policy** entry alongside the built-in sections. The panel follows DSH's language

(Settings → General → Language) and defaults to English; Chinese is included.

Everything there is saved to DSH's normal settings document and applies

immediately — no restart needed. Each field also has a restore default button

that removes your override.

Settings

Field · Default · What it means

Enable · on · Turn the plugin off to get DSH's stock behaviour back.

Routes · · Which routes it governs: provider or provider/model, comma or space separated. = every route.

Trigger ratio · 0.8 · Condense once the context reaches this share of the window. 0.8 = 80%, DSH's own default.

Output reserve (tokens) · 24576 · The margin kept free for the model's answer. DSH's own value is 65536; this is the knob that fixes the early trigger.

Choosing values

General rule first, examples second.

• Trigger ratio — leave it at 0.8 unless you want extra safety. On a very

small window some people prefer 0.7.

• Output reserve — set it to the longest answer you realistically expect, plus

a little slack. 16384–32768 covers most chat and coding work; a model that

writes very long files may want more. Too small a reserve risks a request being

rejected because the answer had nowhere to go.

• Routes — narrow it when you only want some providers affected. For example

llama-cpp (only local models served by llama.cpp), ollama, `o

…(truncated)