fix(web-search): support Exa baseUrl

CHANGELOG.md

@@ -25,6 +25,7 @@ Docs: https://docs.openclaw.ai
 
 - Feishu: preserve Feishu/Lark HTTP error bodies for message sends, media sends, and chat member lookups, so HTTP 400 failures include vendor code, message, log id, and troubleshooter details. Fixes #73860. Thanks @desksk.
 - Agents/transcripts: avoid reopening large Pi transcript files through the synchronous session manager for maintenance rewrites, persisted tool-result truncation, manual compaction boundary hardening, and queued compaction rotation. Thanks @mariozechner.
+- Web search/Exa: accept `plugins.entries.exa.config.webSearch.baseUrl`, normalize it to the Exa `/search` endpoint, and partition cached results by endpoint. Fixes #54928 and supersedes #54939. Thanks @mrpl327 and @lyfuci.
 - Web search/MiniMax: include MiniMax Search in the web-search setup flow and let `MINIMAX_API_KEY` participate in MiniMax Search auto-detection. Supersedes #65828. Thanks @Jah-yee.
 - Telegram: inherit the process DNS result order for Bot API transport and downgrade recovered sticky IPv4 fallback promotions to debug logs, while keeping pinned-IP escalation warnings visible. Fixes #75904. Thanks @highfly-hi and @neeravmakwana.
 - Web search/MiniMax: allow `MINIMAX_OAUTH_TOKEN` to satisfy MiniMax Search credentials, so OAuth-authorized MiniMax Token Plan setups do not need a separate web-search key. Fixes #65768. Thanks @kikibrian and @zhouhe-xydt.

docs/.generated/config-baseline.sha256

@@ -1,4 +1,4 @@
-ae25cb1d397f1ea9642047ef13d35300c807cb1cd67f681c0b5af83b572b3638  config-baseline.json
-0a1907d595765b8bb7a41348d14323920ab50e402be49a19a45a4e2499306407  config-baseline.core.json
-c401cd3450f1737bc92418cfea301d20b54b7fbef9e6049834acc01af338e538  config-baseline.channel.json
-7731a0b93cb335b56fac4c807447ba659fea51ea7a6cd844dc0ef5616669ee75  config-baseline.plugin.json
+051884bad7339a302ecb75e5f61831b1726c6f0360de27485aac76097570c808  config-baseline.json
+80e6e8dce647aef2d1310de55a81d27de52cca47fc24bd7ad81b80f43a72b84c  config-baseline.core.json
+eab8a85eefa2792fb8b98a07698e5ec31ff0b6f8af6222767e8049dcc5c4f529  config-baseline.channel.json
+6bd6c72b17801072b2d3285c82f4c21adcc95f0edffc1e6f64e767d0a07b678f  config-baseline.plugin.json

docs/tools/exa-search.md

@@ -38,6 +38,7 @@ extraction (highlights, text, summaries).
         config: {
           webSearch: {
             apiKey: "exa-...", // optional if EXA_API_KEY is set
+            baseUrl: "https://api.exa.ai", // optional; OpenClaw appends /search
           },
         },
       },
@@ -56,6 +57,14 @@ extraction (highlights, text, summaries).
 **Environment alternative:** set `EXA_API_KEY` in the Gateway environment.
 For a gateway install, put it in `~/.openclaw/.env`.
 
+## Base URL override
+
+Set `plugins.entries.exa.config.webSearch.baseUrl` when Exa search requests
+should go through a compatible proxy or alternate Exa endpoint. OpenClaw
+normalizes bare hosts by prepending `https://` and appends `/search` unless the
+path already ends there. The resolved endpoint is included in the search cache
+key, so results from different Exa endpoints are not shared.
+
 ## Tool parameters
 
 <ParamField path="query" type="string" required>

docs/tools/web.md

@@ -170,7 +170,7 @@ API-backed providers first:
 5. **Kimi** -- `KIMI_API_KEY` / `MOONSHOT_API_KEY` or `plugins.entries.moonshot.config.webSearch.apiKey` (order 40)
 6. **Perplexity** -- `PERPLEXITY_API_KEY` / `OPENROUTER_API_KEY` or `plugins.entries.perplexity.config.webSearch.apiKey` (order 50)
 7. **Firecrawl** -- `FIRECRAWL_API_KEY` or `plugins.entries.firecrawl.config.webSearch.apiKey` (order 60)
-8. **Exa** -- `EXA_API_KEY` or `plugins.entries.exa.config.webSearch.apiKey` (order 65)
+8. **Exa** -- `EXA_API_KEY` or `plugins.entries.exa.config.webSearch.apiKey`; optional `plugins.entries.exa.config.webSearch.baseUrl` overrides the Exa endpoint (order 65)
 9. **Tavily** -- `TAVILY_API_KEY` or `plugins.entries.tavily.config.webSearch.apiKey` (order 70)
 
 Key-free fallbacks after that:

extensions/exa/openclaw.plugin.json

@@ -12,6 +12,10 @@
       "help": "Exa Search API key (fallback: EXA_API_KEY env var).",
       "sensitive": true,
       "placeholder": "exa-..."
+    },
+    "webSearch.baseUrl": {
+      "label": "Exa Search Base URL",
+      "help": "Optional Exa Search API base URL override. OpenClaw appends /search when the URL does not already end there."
     }
   },
   "contracts": {
@@ -30,6 +34,9 @@
         "properties": {
           "apiKey": {
             "type": ["string", "object"]
+          },
+          "baseUrl": {
+            "type": "string"
           }
         }
       }

extensions/exa/src/exa-web-search-provider.runtime.ts

@@ -29,6 +29,7 @@ const EXA_MAX_SEARCH_COUNT = 100;
 
 type ExaConfig = {
   apiKey?: string;
+  baseUrl?: string;
 };
 
 type ExaSearchType = (typeof EXA_SEARCH_TYPES)[number];
@@ -87,6 +88,44 @@ function resolveExaApiKey(exa?: ExaConfig): string | undefined {
   );
 }
 
+function invalidBaseUrlPayload(value: string) {
+  return {
+    error: "invalid_base_url",
+    message: `plugins.entries.exa.config.webSearch.baseUrl must be a valid http(s) URL. Got: ${value}`,
+    docs: "https://docs.openclaw.ai/tools/exa-search",
+  };
+}
+
+function resolveExaSearchEndpoint(
+  exa?: ExaConfig,
+): { endpoint: string } | { error: string; message: string; docs: string } {
+  const configured = normalizeOptionalString(exa?.baseUrl);
+  if (!configured) {
+    return { endpoint: EXA_SEARCH_ENDPOINT };
+  }
+
+  if (/^[a-z][a-z0-9+.-]*:\/\//i.test(configured) && !/^https?:\/\//i.test(configured)) {
+    return invalidBaseUrlPayload(configured);
+  }
+  const candidate = /^https?:\/\//i.test(configured) ? configured : `https://${configured}`;
+  let parsed: URL;
+  try {
+    parsed = new URL(candidate);
+  } catch {
+    return invalidBaseUrlPayload(configured);
+  }
+  if (parsed.protocol !== "http:" && parsed.protocol !== "https:") {
+    return invalidBaseUrlPayload(configured);
+  }
+
+  const pathname = parsed.pathname.replace(/\/+$/, "");
+  parsed.pathname = pathname.endsWith("/search")
+    ? pathname
+    : `${pathname === "" ? "" : pathname}/search`;
+  parsed.hash = "";
+  return { endpoint: parsed.toString() };
+}
+
 function resolveExaDescription(result: ExaSearchResult): string {
   const highlights = result.highlights;
   if (Array.isArray(highlights)) {
@@ -315,6 +354,7 @@ function resolveFreshnessStartDate(freshness: ExaFreshness): string {
 
 async function runExaSearch(params: {
   apiKey: string;
+  endpoint: string;
   query: string;
   count: number;
   freshness?: ExaFreshness;
@@ -342,7 +382,7 @@ async function runExaSearch(params: {
 
   return withTrustedWebSearchEndpoint(
     {
-      url: EXA_SEARCH_ENDPOINT,
+      url: params.endpoint,
       timeoutSeconds: params.timeoutSeconds,
       init: {
         method: "POST",
@@ -378,6 +418,31 @@ function missingExaKeyPayload() {
   };
 }
 
+function buildExaCacheKey(params: {
+  endpoint: string;
+  type: ExaSearchType;
+  query: string;
+  count: number;
+  freshness?: ExaFreshness;
+  dateAfter?: string;
+  dateBefore?: string;
+  contents?: ExaContentsArgs;
+}): string {
+  return buildSearchCacheKey([
+    "exa",
+    params.endpoint,
+    params.type,
+    params.query,
+    params.count,
+    params.freshness,
+    params.dateAfter,
+    params.dateBefore,
+    params.contents?.highlights ? JSON.stringify(params.contents.highlights) : undefined,
+    params.contents?.text ? JSON.stringify(params.contents.text) : undefined,
+    params.contents?.summary ? JSON.stringify(params.contents.summary) : undefined,
+  ]);
+}
+
 export async function executeExaWebSearchProviderTool(
   ctx: { config?: Record<string, unknown>; searchConfig?: SearchConfigRecord },
   args: Record<string, unknown>,
@@ -393,6 +458,11 @@ export async function executeExaWebSearchProviderTool(
   if (!apiKey) {
     return missingExaKeyPayload();
   }
+  const endpointResult = resolveExaSearchEndpoint(exaConfig);
+  if ("error" in endpointResult) {
+    return endpointResult;
+  }
+  const endpoint = endpointResult.endpoint;
 
   const query = readStringParam(params, "query", { required: true });
   const rawType = readStringParam(params, "type");
@@ -442,18 +512,17 @@ export async function executeExaWebSearchProviderTool(
       ? parsedContents.value
       : undefined;
 
-  const cacheKey = buildSearchCacheKey([
-    "exa",
+  const resolvedCount = resolveExaSearchCount(count, DEFAULT_SEARCH_COUNT);
+  const cacheKey = buildExaCacheKey({
+    endpoint,
     type,
     query,
-    resolveExaSearchCount(count, DEFAULT_SEARCH_COUNT),
+    count: resolvedCount,
     freshness,
     dateAfter,
     dateBefore,
-    contents?.highlights ? JSON.stringify(contents.highlights) : undefined,
-    contents?.text ? JSON.stringify(contents.text) : undefined,
-    contents?.summary ? JSON.stringify(contents.summary) : undefined,
-  ]);
+    contents,
+  });
   const cached = readCachedSearchPayload(cacheKey);
   if (cached) {
     return cached;
@@ -462,8 +531,9 @@ export async function executeExaWebSearchProviderTool(
   const start = Date.now();
   const results = await runExaSearch({
     apiKey,
+    endpoint,
     query,
-    count: resolveExaSearchCount(count, DEFAULT_SEARCH_COUNT),
+    count: resolvedCount,
     freshness,
     dateAfter,
     dateBefore,
@@ -519,9 +589,11 @@ export const __testing = {
   normalizeExaResults,
   normalizeExaFreshness,
   parseExaContents,
+  buildExaCacheKey,
   resolveExaApiKey,
   resolveExaConfig,
   resolveExaDescription,
   resolveExaSearchCount,
+  resolveExaSearchEndpoint,
   resolveFreshnessStartDate,
 } as const;

extensions/exa/src/exa-web-search-provider.test.ts

@@ -46,6 +46,40 @@ describe("exa web search provider", () => {
     expect(__testing.resolveExaApiKey({ apiKey: "exa-secret" })).toBe("exa-secret");
   });
 
+  it("resolves Exa search base URL overrides", () => {
+    expect(__testing.resolveExaSearchEndpoint()).toEqual({
+      endpoint: "https://api.exa.ai/search",
+    });
+    expect(__testing.resolveExaSearchEndpoint({ baseUrl: "https://proxy.example/exa" })).toEqual({
+      endpoint: "https://proxy.example/exa/search",
+    });
+    expect(__testing.resolveExaSearchEndpoint({ baseUrl: "proxy.example/exa/search/" })).toEqual({
+      endpoint: "https://proxy.example/exa/search",
+    });
+    expect(__testing.resolveExaSearchEndpoint({ baseUrl: "ftp://proxy.example/exa" })).toEqual(
+      expect.objectContaining({ error: "invalid_base_url" }),
+    );
+  });
+
+  it("partitions Exa cache keys by resolved endpoint", () => {
+    const base = {
+      type: "auto" as const,
+      query: "openclaw",
+      count: 5,
+    };
+    expect(
+      __testing.buildExaCacheKey({
+        ...base,
+        endpoint: "https://api.exa.ai/search",
+      }),
+    ).not.toBe(
+      __testing.buildExaCacheKey({
+        ...base,
+        endpoint: "https://proxy.example/exa/search",
+      }),
+    );
+  });
+
   it("normalizes Exa result descriptions from highlights before text", () => {
     expect(
       __testing.resolveExaDescription({