* fix(#2615): document the effortSurface axis in the host-integration matrix #2481 added `effortSurface` as the ninth negotiated `hostIntegration` axis and wrote documentation-sourced values into 18 descriptors, but never touched `docs/reference/host-integration-capability-matrix.md`. The matrix that ADR-1239 designates the cited source of truth had zero occurrences of the axis: no entry in the axes legend, and no row in any of the per-runtime tables. `src/host-integration.cts` states "every value is documented or explicitly 'undocumented'" — for this axis that was false for every runtime. Adds the legend entry (the `argv` / `none` / `undocumented` vocabulary, plus why there is deliberately no config-file member) and an `effortSurface` row to all 19 per-runtime tables. Every citation is carried over from #2481's own commit message, where the values were sourced: - claude argv -- `claude --help` documents `--effort <level>` - opencode argv -- `opencode run --help` documents `--variant` - codex argv -- `model_reasoning_effort` is a config.toml key, not a dedicated flag, so the generic `-c key=value` override is the only argv route (still argv) - 15 hosts undocumented -- their docs state no reasoning setting kimi-code is the nineteenth section (added by #2603 after #2481) and is the one runtime with no declared value. Its row and a Documentation-gaps entry record why rather than inventing one: Kimi Code documents `/effort` (alias `/thinking`), but only as an INTERACTIVE slash command — `-m, --model` is the only model-adjacent argv. Neither vocabulary member is accurate (`none` would deny a mechanism the host has, `argv` would claim one it does not expose), so closing that gap needs a vocabulary decision, which is a negotiation change and not a documentation one. The absent value already degrades closed exactly as the sentinel does. The regression test derives its runtime list from the registry rather than hardcoding it, so a runtime added later fails until its matrix row exists — the ratchet whose absence let #2481 add an axis with nothing catching the missing docs. Closes #2615 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QT3ibz5qJuDuGqpTGRYVGf * docs(#2615): honest citations for the undocumented rows; fix the four stale 8-axis lists Two findings from the orthogonal review of the first commit. 1. The 15 `undocumented` rows shared byte-identical text — "searched the runtime's official docs (see Sources consulted above)" — which is weaker than this file's own convention ("no authoritative doc — searched: <url>") and, worse, implies a per-host targeted search that did not happen: each section's Sources-consulted list was gathered for OTHER axes and contains no CLI-reference or reasoning-effort source. The rows now say plainly what the finding is — an ABSENCE established by #2481's cross-host survey — and cite that survey rather than implying a URL was checked per host. 2. Four normative docs still described "the eight negotiated axes" and omitted effortSurface entirely. The worst of them is docs/how-to/add-or-update-a-host-integration.md — the maintainer's own guide for onboarding a host, whose Step 2 axis table would have a maintainer reproduce exactly the gap #2615 exists to close. Also fixed: docs/reference/host-integration-interface.md (which calls itself the normative reference and had no effortSurface row at all), docs/how-to/author-a-host-plugin.md, docs/registries/README.md ("**exactly** the eight … axes keys"), and CONTEXT.md's matching EoS-registry sentence. Deliberately NOT changed, because they are historical records rather than current contract: docs/whats-new-1.7.0.md and docs/FEATURES.md's 1.7.0 entry (effortSurface shipped in 1.8.0 via #2481 — rewriting them would falsify the release history), ADR-1239's pre-amendment body (already superseded by its own "Amendment (2026-07-21): effortSurface axis (#2481)"), and ADR-1016's "original eight axes", which refers to a different axis set entirely. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QT3ibz5qJuDuGqpTGRYVGf * chore(#2615): backfill changeset PR number (#2698) --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
3.4 KiB
How to author a host-plugin for GSD
This guide is for external tool authors who want to embed GSD's orchestration loop (dispatch + hooks + model + state) into a new host — a CLI, IDE, or agent runtime GSD does not special-case — by writing a host-plugin against the published Host-Integration SDK, without modifying gsd-core source.
The contract is the SDK entry (src/host-integration-sdk.cts →
gsd-core/bin/lib/host-integration-sdk.cjs): import only from it. Everything it
re-exports is public and versioned (PROTOCOL_VERSION); everything else in
gsd-core is internal and may change. A smoke test
(tests/sdk-smoke.test.cjs) proves a third-party host can be built from this
surface alone.
1. Decide your host's profile
Classify your host into one of three profiles (the SDK's profileOf does this
from your axes):
| Profile | Example hosts | Binding |
|---|---|---|
programmatic-cli |
OpenCode, pi | imperative adapter (createImperativeAdapter) |
declarative-cli |
Antigravity, Codex | declarative adapter (createDeclarativeAdapter) |
ide |
VS Code | imperative adapter + engine-owned hook bus + active model + sandboxed state |
2. Declare your host's integration axes
Declare the nine negotiated axes (embeddingMode, commandSurface, dispatch,
modelMode, hookBus, stateIO, transport, runtime, effortSurface) from your host's
authoritative documentation — never infer. Where the docs are silent, use the
undocumented sentinel (the SDK degrades it fail-closed). See
docs/reference/host-integration-interface.md for the closed vocabulary.
3. Compose the engine adapters
const SDK = require('@opengsd/gsd-core/sdk'); // the published entry
// Engine-owned hook bus (for hosts with no event bus, e.g. VS Code):
const hookBus = SDK.createHookBus({ bus: 'engine' });
// Active model via a host provider (e.g. vscode.lm); or 'passive' for CLIs:
const model = SDK.createModelAdapter({ modelMode: 'active' }, { sendRequest: hostSendRequest });
// State IO — filesystem for CLIs, sandboxed-storage for IDEs/web:
const stateIO = SDK.createStateIO({ io: 'filesystem' });
// The engine-as-library adapter bound to your runtime id:
const adapter = SDK.createImperativeAdapter({ runtime: 'my-host' });
4. Negotiate capabilities (serialized, for out-of-process hosts)
If your host runs out-of-process (can't share object refs with the engine), use
the serialized handshake instead of the in-process negotiateHostCapabilities:
const req = SDK.buildHandshakeRequest({ ...myAxes });
const { effective, points, warnings } = SDK.handleHandshakeRequest(req);
The result is identical to the in-process negotiation for the same axes — that consistency is the contract.
5. Bind your host primitives + ship
Map the six interface points (command, dispatch, model, hooks, state, artifact) to your host's primitives (palette/chat, plugin API, provider calls, etc.). The reference host-plugins (OpenCode, Antigravity, pi, VS Code — see Phase 5) are canonical examples to mirror. Your host-plugin is now a self-contained artifact; no gsd-core source change is required.
See also
- Tutorial:
docs/tutorials/embed-gsd-in-a-new-host.md— a complete end-to-end embedding. - Reference:
docs/reference/host-integration-interface.md— every axis, adapter, and the handshake. - Versioning:
docs/explanation/interface-versioning-policy.md— what evolves additively vs. breaking.