test: cover OpenAI server compaction docs

docs/gateway/configuration-reference.md

@@ -1235,9 +1235,10 @@ Time format in system prompt. Default: `auto` (OS preference).
 - `verboseDefault`: default verbose level for agents. Values: `"off"`, `"on"`, `"full"`. Default: `"off"`.
 - `elevatedDefault`: default elevated-output level for agents. Values: `"off"`, `"on"`, `"ask"`, `"full"`. Default: `"on"`.
 - `model.primary`: format `provider/model` (e.g. `openai/gpt-5.4` for API-key access or `openai-codex/gpt-5.5` for Codex OAuth). If you omit the provider, OpenClaw tries an alias first, then a unique configured-provider match for that exact model id, and only then falls back to the configured default provider (deprecated compatibility behavior, so prefer explicit `provider/model`). If that provider no longer exposes the configured default model, OpenClaw falls back to the first configured provider/model instead of surfacing a stale removed-provider default.
-- `models`: the configured model catalog and allowlist for `/model`. Each entry can include `alias` (shortcut) and `params` (provider-specific, for example `temperature`, `maxTokens`, `cacheRetention`, `context1m`).
+- `models`: the configured model catalog and allowlist for `/model`. Each entry can include `alias` (shortcut) and `params` (provider-specific, for example `temperature`, `maxTokens`, `cacheRetention`, `context1m`, `responsesServerCompaction`, `responsesCompactThreshold`).
   - Safe edits: use `openclaw config set agents.defaults.models '<json>' --strict-json --merge` to add entries. `config set` refuses replacements that would remove existing allowlist entries unless you pass `--replace`.
   - Provider-scoped configure/onboarding flows merge selected provider models into this map and preserve unrelated providers already configured.
+  - For direct OpenAI Responses models, server-side compaction is enabled automatically. Use `params.responsesServerCompaction: false` to stop injecting `context_management`, or `params.responsesCompactThreshold` to override the threshold. See [OpenAI server-side compaction](/providers/openai#server-side-compaction-responses-api).
 - `params`: global default provider parameters applied to all models. Set at `agents.defaults.params` (e.g. `{ cacheRetention: "long" }`).
 - `params` merge precedence (config): `agents.defaults.params` (global base) is overridden by `agents.defaults.models["provider/model"].params` (per-model), then `agents.list[].params` (matching agent id) overrides by key. See [Prompt Caching](/reference/prompt-caching) for details.
 - `embeddedHarness`: default low-level embedded agent runtime policy. Use `runtime: "auto"` to let registered plugin harnesses claim supported models, `runtime: "pi"` to force the built-in PI harness, or a registered harness id such as `runtime: "codex"`. Set `fallback: "none"` to disable automatic PI fallback.

docs/providers/openai.md

@@ -661,12 +661,14 @@ the Server-side compaction accordion below.
   </Accordion>
 
   <Accordion title="Server-side compaction (Responses API)">
-    For direct OpenAI Responses models (`openai/*` on `api.openai.com`), OpenClaw auto-enables server-side compaction:
+    For direct OpenAI Responses models (`openai/*` on `api.openai.com`), the OpenAI plugin's Pi-harness stream wrapper auto-enables server-side compaction:
 
     - Forces `store: true` (unless model compat sets `supportsStore: false`)
     - Injects `context_management: [{ type: "compaction", compact_threshold: ... }]`
     - Default `compact_threshold`: 70% of `contextWindow` (or `80000` when unavailable)
 
+    This applies to the built-in Pi harness path and to OpenAI provider hooks used by embedded runs. The native Codex app-server harness manages its own context through Codex and is configured separately with `agents.defaults.embeddedHarness.runtime`.
+
     <Tabs>
       <Tab title="Enable explicitly">
         Useful for compatible endpoints like Azure OpenAI Responses:

extensions/openai/openai-provider.test.ts

@@ -447,6 +447,7 @@ describe("buildOpenAIProvider", () => {
         provider: "openai",
         id: "gpt-5.4",
         baseUrl: "https://api.openai.com/v1",
+        contextWindow: 200_000,
       } as Model<"openai-responses">,
       payload: {
         reasoning: { effort: "none" },
@@ -457,6 +458,10 @@ describe("buildOpenAIProvider", () => {
       transport: "auto",
       openaiWsWarmup: true,
     });
+    expect(result.payload.store).toBe(true);
+    expect(result.payload.context_management).toEqual([
+      { type: "compaction", compact_threshold: 140_000 },
+    ]);
     expect(result.payload.service_tier).toBe("priority");
     expect(result.payload.text).toEqual({ verbosity: "low" });
     expect(result.payload.reasoning).toEqual({ effort: "none" });

src/agents/openai-responses-payload-policy.test.ts

@@ -32,6 +32,33 @@ describe("openai responses payload policy", () => {
     });
   });
 
+  it("couples native Responses server compaction to provider-managed store", () => {
+    const model = {
+      id: "gpt-5.4",
+      api: "openai-responses",
+      provider: "openai",
+      baseUrl: "https://api.openai.com/v1",
+      contextWindow: 200_000,
+    } satisfies Pick<
+      Model<"openai-responses">,
+      "api" | "baseUrl" | "contextWindow" | "id" | "provider"
+    >;
+    const payload = {} satisfies Record<string, unknown>;
+
+    applyOpenAIResponsesPayloadPolicy(
+      payload,
+      resolveOpenAIResponsesPayloadPolicy(model, {
+        enableServerCompaction: true,
+        storeMode: "provider-policy",
+      }),
+    );
+
+    expect(payload).toEqual({
+      store: true,
+      context_management: [{ type: "compaction", compact_threshold: 140_000 }],
+    });
+  });
+
   it("strips store and prompt cache for proxy-like responses routes when requested", () => {
     const policy = resolveOpenAIResponsesPayloadPolicy(
       {