@local/laya-tool-router
延迟工具发现：从预填充中隐藏所有工具模式，并按需通过 Laya Choice 显示其中一个。
插件会安装到这里；不确定时保持 web。
npx -y @deepseek-ai/dsh plugin --profile web add github:residenthiago2011/DSH-Laya-Tool-Router#a273e6a974ff487f06eb4b1e06c00b1d53f94bffLaya Tool Router 以 @local/laya-tool-router 发布，当前版本为 1.0.0。Plugin Hub 会校验它的 manifest，并保存精确安装来源，便于复现安装结果。
Lazy tool discovery for DSH: hides every tool schema from the model's prefill and reveals exactly one tool on demand through the Laya decision endpoint.
find_tool schema instead of every mounted tool
(26 schemas on a default Standard profile — see Why?).POST /v1/systemone to Laya: a bestTool choice plus a
usable noul question (~150 tokens).confidence >= 0.30 and p(noul) <= 0.30: return
the winner's JSON Schema and reveal exactly that one tool; otherwise return
a shortlist. Exact tool-name requests skip Laya and reveal instantly.Fresh session, first message, server-reported usage.prompt_tokens —
the DSH Standard preset (the platform default) versus this plugin's
no-prefill preset:
| Cost before the first word | Standard | no-prefill |
|---|---|---|
| Prefill tokens | 6,559 | 283 |
| System prompt | 5,816 chars | 372 chars |
| Tool schemas shipped up front | 26 schemas (21,417 B) | 1 schema (697 B) |
Standard makes the model ingest a 5.8k-character system prompt plus every
mounted tool schema — 6,559 tokens — before it emits a single word, and pays
it again on every subsequent turn. no-prefill ingests 283 tokens (23×
less, 96% smaller prefill): a slim persona plus the single find_tool
schema. On the measured backend that difference was ~40 s versus ~14 min of
prompt evaluation for the first reply.
What that buys you:
If a session ends up using most tools anyway, the gap narrows — but the unused half of the shelf is exactly what a default profile charges for from word one.
cordis.patch.yml (applied when the bundle is selected):
preset-no-prefill — an agent preset whose composition is only
persona + laya-tool-router. The persona tells the model: no tools are
declared; call find_tool first.tool-bash, tool-fs, tool-fs-search,
tool-todo, tool-web flipped to disabled: false, plus a present
row inserted. Discovery needs the tools to exist somewhere; preset-scoped
registrations shadow host ones, so other presets keep their own behavior.
The slim preset keeps the Agent Presets page showing only find_tool.dsh.bundle.patch).http://127.0.0.1:8081/v1/systemone
(edit LAYA_BASE in index.js to change; default port 8081).pnpm on PATH of the host running DSH (the plugin manager installs
through pnpm; the CLI forwards pnpm arguments).From GitHub (recommended): Plugins → install → spec
git+https://github.com/residenthiago2011/DSH-Laya-Tool-Router.git
or via the CLI:
dsh plugin --profile <profile> add git+https://github.com/residenthiago2011/DSH-Laya-Tool-Router.git
From a prebuilt tarball: download
laya-tool-router.tgz
from Releases, then install from the local path (also accepted: absolute
path, git URL, or an npm registry name after publishing):
dsh plugin --profile <profile> add /path/to/laya-tool-router.tgz
Then select the bundle on the Plugins page (installation enables it by default) and restart DSH if the profile runs without HMR.
Deselect/remove the bundle on the Plugins page. The preset row and the host tool flips vanish with the bundle patch; other presets are untouched.
| File | Purpose |
|---|---|
index.js | The router plugin: registers find_tool, masks every other tool per agent, runs the 3-stage decision. |
cordis.patch.yml | The bundle patch: preset declaration + host tool enablement. |
tool-registry.json | Stage-0 labels (25 tools, including bash) and Stage-1 short criteria. |
node router-import-test.mjs # module shape (no dependencies)
node rank-sim.mjs # Stage-0 ranking on sample queries (needs tool-registry.json)
node router-e2e-test.mjs # full flow: arm/deny-set/live reveal — requires Laya at 127.0.0.1:8081
router-e2e-test.mjs uses a stub DSH context that emits tools/change
synchronously (the worst case that once caused a ~970x feedback loop), then
drives the real find_tool against a live Laya endpoint: armed-once,
deny-set dedup, late registration, reveal re-arm, empty-request catalog hint,
and the exact-name fast path. ALL_PASS is expected.
find_tool registers with a standard JSON Schema
(type/properties/required/additionalProperties). A raw register with
author-style parameters reaches the LLM unconstrained — the model then
calls with {}. Always compile.execute() in index.js.