dsh-terse
Mandatory, always-on terseness for DeepSeek Harness
agents. Compresses everything an agent emits and everything it reads, without replacing the
deployment's system prompt and without giving up answer quality.
There is no mode and no off switch. Mounting the row is the enabling act.
What it does
| Layer | Extension point | Effect |
|---|
| L1 constitution | systemPrompt.section | A static, cache-stable rule set that makes every reply compressed. |
| L2 standing reminder | systemPrompt.context | A durable <system-reminder> snapshot that survives compaction, where a system-only rule fades. |
| L3 drift nudge | tools/post-execute -> additionalContexts | A one-line reminder that rides a tool result and catches a reply that ran long. Capped per turn; cannot loop. |
| L4 input shaping | tools/post-execute content | Collapses repeated lines and elides the middle of oversized tool output, with a locator so the original is one call away. |
| L5 enforcement | system-prompt/assemble | Asserts the constitution survived assembly; logs a loud warning if it did not. Never rewrites peer contributions. |
| L6 coverage | per-session state | Cadence and reply observations keyed by the session object, so a resumed or forked session starts clean. |
Why it does not hurt quality
Compression is aimed at the chat mouth, never at meaning or at durable artifacts:
- Never cut negations, quantifiers, numbers, units, code, identifiers, paths, commands, or
exact error strings. These are the token classes whose loss flips semantics.
- Never fake terseness. Invented abbreviations and arrows cost the same tokens and hurt
clarity; if the terse form is not shorter than plain, plain wins.
- Full prose returns for security warnings, irreversible-action confirmations,
order-sensitive multi-step sequences, and anywhere compression would create ambiguity.
- Code is never judged as drift, so a long implementation is never punished.
- Tool output is only elided reversibly, never lossy on errors, and the locator lets the
agent pull the original back.
Those choices are measured in benchmark/README.md; their design record
is a local working artifact, not a shipped doc.
Install
The plugin is one row, and it peers on the harness window >=0.1.5-rc.1 <0.3.0; the
releases verified for it are listed in dsh.compatibility.dshReleases of
package.json. Add it to a profile:
dsh plugin --profile <name> add @sagmans/dsh-terse
Or insert it directly in the profile's cordis.patch.yml:
- insert:
- id: terse
name: '@sagmans/dsh-terse'
The shipped row is in cordis.patch.yml.
Configuration
Every field is optional; the shipped defaults are the product.
| Field | Default | Meaning |
|---|
constitution | shipped text | Replaces the L1 rule set. |
standing | shipped text | Replaces the L2 reminder; empty suppresses it. |
shaping | true | Whether L4 runs. |
drift | see below | L3 thresholds. |
shapingConfig | see below | L4 bounds. |
interface DriftConfig {
maxReplyTokens: number // 220 reply length above which the prose budget is exceeded
minReplyTokensToJudge: number // 60 floor below which no reply is judged
maxProseRatio: number // 0.72 prose ratio above which a long reply is padding
minJudgeableProseRatio: number // 0.4 below this the reply is code, never judged
idleStepsBeforeNudge: number // 14 steps with no nudge before the anti-drift backstop fires
maxNudgesPerTurn: number // 2 hard cap on nudges, and so on their extra steps
}
interface ShapingConfig {
maxResultChars: number // 12000 threshold above which a result may be elided
headChars: number // 3000 head kept
tailChars: number // 1500 tail kept, where error trails live
duplicateRunMin: number // 4 repeated lines before a run collapses
}
Development
pnpm install --frozen-lockfile
pnpm run typecheck
pnpm test
pnpm run build
A profile loads dist/, not src/; run pnpm run build before dogfooding a linked profile.
Try it in the TUI
scripts/dogfood/run-terse-from-worktree.sh clones the real ~/.dsh to a scratch
directory, mounts this checkout on the real tui profile inside the clone, and launches
it. Every composition input is present — your other bundles, the profile patch, settings,
themes, credentials — while every write goes to the throwaway clone.
./scripts/dogfood/run-terse-from-worktree.sh # clone, build, launch
./scripts/dogfood/run-terse-from-worktree.sh --no-launch # set up, print the command
./scripts/dogfood/run-terse-from-worktree.sh --status # show what the clone runs
./scripts/dogfood/run-terse-from-worktree.sh --with-sessions
./scripts/dogfood/run-terse-from-worktree.sh --clean # remove the clone
What to look for on the surface: the status line shows injected … terse:standing … when
the layer is live. Ask a short factual question (expect one line), ask for code inline
(expect the full implementation, no elision), and ask something safety-critical (expect
full prose). Those three cases are the contract; if the code case is trimmed, that is a bug.
Benchmark
benchmark/benchmark.sh measures this plugin against a control profile that lacks it. See
benchmark/README.md; the neutral home is mandatory, because a
profile whose global instructions already ask for brevity makes the control terse too.
License
MIT.
Git hooks
Checks run in tracked local Git hooks instead of push/PR CI. Run scripts/setup-hooks once after cloning; it points this clone's core.hooksPath at .githooks. Each hook runs the generic checks first (sign-off and commit message format, a gitleaks secret scan, and signed, signed-off commits on push), then this repository's checks in .githooks/<hook>.d/ in name order. Install gitleaks first. Every check must pass before you push; never use --no-verify. Pre-push checks test exactly each pushed commit: in place only when it is the clean, unmodified checkout, otherwise in a temporary worktree of that commit.
This repository's checks:
pre-commit: staged whitespace and conflict-marker check.pre-push: npm audit signatures, pnpm run check, pnpm audit --audit-level high.
