john-walks-slow/dsh-set-model

A DeepSeek Harness (DSH / cordis) plugin for dynamic model and reasoning-effort switching: the agent autonomously changes its provider, model and reasoningEffort via the set_model / list_models tools based on task difficulty, the plugin injects the active model into the runtime context, and models are upgraded/downgraded automatically alongside Plan Mode.

Switch to deep reasoning for hard problems, stay on the cost-effective tier for routine work — no manual configuration changes needed.

What the model sees

Via systemPrompt.context, the plugin appends one line with the current active model to the tail of every step's Runtime Context snapshot (cache-safe append; the model can check itself at any time without calling a tool):

[Current active model: deepseek/deepseek-chat · reasoning: low]

While Plan Mode is active, a status marker is attached:

[Current active model: deepseek/deepseek-reasoner · reasoning: high] [Plan Mode Active]

Tool: set_model

set_model({ provider?, model?, reasoningEffort?, reason }) — reason is mandatory (every switch invalidates the session's KV cache, so the tool description requires the agent to justify the switch and avoid frequent changes):

{ "model": "deepseek-reasoner", "reasoningEffort": "high", "reason": "Hard concurrency debugging, deep reasoning required" }

Tool output:

Model successfully switched to deepseek/deepseek-reasoner · reasoning: high. Effective from the next step.
Reason: Hard concurrency debugging, deep reasoning required

Tool: list_models

list_models({ provider? }) lists registered providers and model capabilities (reasoning efforts, context window):

# Available Providers & Models:

## Provider: deepseek
- **deepseek-reasoner** (DeepSeek Reasoner) | Reasoning: [low, high, max] | Context: 128000
- **deepseek-chat** (DeepSeek Chat) | Context: 128000

A switch takes effect from the next step: the plugin appends a model/selection event to the session (the same mechanism as the GUI model picker), and the Web UI state stays in sync.

Plan Mode auto-transition

The plugin watches the plan projection state on agent/pre-step — zero manual intervention:

• Entering Plan Mode (/plan on): stashes the current everyday model and automatically switches to the configured planModel (a high-reasoning planning model);
• Leaving Plan Mode (exit_plan_mode approval / /plan off): automatically restores the pre-entry model (autoRestorePlanModel, on by default);
• A failed auto-switch only degrades to a warn and never interrupts the step.

Safety guardrails

• Provider whitelist: when allowedProviders is non-empty, switching to providers outside the whitelist is rejected outright;

• Context-window guard: if the current session's token pressure exceeds the target model's context window, the switch is rejected with a prompt to compact context first (e.g. clear_mind):

  Current session token pressure (70000 tokens) exceeds target model context window (65536 tokens).
  Please compact context (e.g. clear_mind) before switching to this model.
  

• Route validation: the candidate combination is validated by the DSH core via ctx.llm.resolveCallConfig (provider adapter exists, model is valid, reasoningEffort is supported); invalid combinations return a clear error and write no session state;

• Root agents only: the tools are registered for root agents only (subagents get no self-switching tools); set enableAgentTools: false to disable them entirely.

Configuration

Settings namespace dsh-set-model (configurable in the Web Settings UI; changes take effect at runtime):

dsh-set-model:
  planModel:                    # optional: target model applied automatically on entering Plan Mode
    provider: deepseek
    model: deepseek-reasoner
    reasoningEffort: high
  autoRestorePlanModel: true    # restore the everyday model when leaving Plan Mode (default true)
  enableAgentTools: true        # whether to register the set_model / list_models tools (default true)
  allowedProviders: []          # optional: whitelist of switchable providers (empty = unrestricted)

Install

dsh plugin --profile web add dsh-set-model

No manual configuration needed after install — the bundled cordis.patch.yml mounts automatically.

Install straight from GitHub (source install; pnpm ≥10 requires allowing the build script):

dsh plugin --profile web add github:john-walks-slow/dsh-set-model
# The first add is blocked by pnpm: add the package name pnpm prints to
# allowBuilds in ~/.dsh/profiles/web/pnpm-workspace.yaml, then re-run

Permissions & compatibility

• No direct side effects: no external services, no network requests from the plugin itself (model routing and capability queries all go through the DSH core llm service), no direct filesystem writes; it only appends a model/selection event via the session event API (persisted by the DSH session system) and injects one Runtime Context line
• Dependencies: @deepseek-ai/cordis 4.0.2 / @deepseek-ai/dsh-agent, dsh-llm, dsh-session, dsh-tools 0.1.2-rc.1 (aligned with dsh 0.1.2-rc.1 locked versions), zod ^3.24.2; Node ≥ 22.5
• Failure degradation: a failed Plan Mode auto-switch only warns and never breaks the step; tool validation failures return a clear error to the agent without polluting session state

Local development

npm install
npm run build     # tsc → dist/src
npm test          # tsc(incl. test) + node --test dist/test/*.test.js

License

MIT

Release a new version

One command runs tests, bumps the version and packs (npm version also commits and tags):

npm run release        # patch; for bigger changes: npm version minor or major

Then publish with the fingerprint flow and push:

node ~/.agents/skills/npm-publish/scripts/publish-webauthn.cjs /tmp/dsh-set-model-<newver>.tgz
git push --follow-tags

Verify with npm view dsh-set-model version. When releasing several packages, check "do not challenge for the next 5 minutes" on the webauthn page to publish them all with one fingerprint.