diff --git a/CHANGELOG.md b/CHANGELOG.md
index 42904fcc1ef..3debcbe4e93 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -16,6 +16,7 @@ Docs: https://docs.openclaw.ai
 - Browser control: scope standalone loopback auth to the resolved active gateway credential and fail closed when password mode lacks a resolved password, so inactive tokens or passwords no longer authorize browser routes. Fixes #65626. (#65639) Thanks @coygeek.
 - Control UI/Codex harness: emit native Codex app-server assistant and lifecycle completion events so live webchat runs stop spinning without needing a transcript reload fallback. (#70815) Thanks @lesaai.
 - Agents/sessions: persist the runtime-resolved context budget from embedded agent runs, so Codex GPT-5.5 sessions keep the catalog/runtime context cap instead of falling back to the generic 200k status value. Fixes #71294. Thanks @tud0r.
+- Agents/tools: fail runs before model submission when explicit tool allowlists resolve to no callable tools, preventing text-only hallucinated tool results for missing tools such as plugin commands that were not registered. Fixes #71292.
 - Discord/replies: run `message_sending` plugin hooks for Discord reply delivery, including DM targets, so plugins can transform or cancel outbound Discord replies consistently with other channels. Fixes #59350. (#71094) Thanks @wei840222.
 - Control UI/commands: carry provider-owned thinking option ids/labels in session rows and defaults so fresh sessions show and accept dynamic modes such as `adaptive`, `xhigh`, and `max`. Fixes #71269. Thanks @Young-Khalil.
 - Image generation: make explicit `model=` overrides exact-only so failed `openai/gpt-image-2` requests no longer fall through to Gemini or other configured providers, and update `image_generate list` to mention OpenAI Codex OAuth as valid auth for `openai/gpt-image-2`. Fixes #71290 and #71231. Thanks @Young-Khalil and @steipete.
diff --git a/docs/tools/index.md b/docs/tools/index.md
index 07f53cc1896..16c154fe328 100644
--- a/docs/tools/index.md
+++ b/docs/tools/index.md
@@ -125,6 +125,12 @@ config. Deny always wins over allow.
 }
 ```
 
+OpenClaw fails closed when an explicit allowlist resolves to no callable tools.
+For example, `tools.allow: ["query_db"]` only works if a loaded plugin actually
+registers `query_db`. If no built-in, plugin, or bundled MCP tool matches the
+allowlist, the run stops before the model call instead of continuing as a
+text-only run that could hallucinate tool results.
+
 ### Tool profiles
 
 `tools.profile` sets a base allowlist before `allow`/`deny` is applied.
diff --git a/docs/tools/multi-agent-sandbox-tools.md b/docs/tools/multi-agent-sandbox-tools.md
index 1995fab656d..a3711ed327e 100644
--- a/docs/tools/multi-agent-sandbox-tools.md
+++ b/docs/tools/multi-agent-sandbox-tools.md
@@ -207,6 +207,12 @@ If `agents.list[].tools.sandbox.tools` is set, it replaces `tools.sandbox.tools`
 If `agents.list[].tools.profile` is set, it overrides `tools.profile` for that agent.
 Provider tool keys accept either `provider` (e.g. `google-antigravity`) or `provider/model` (e.g. `openai/gpt-5.4`).
 
+If any explicit allowlist in that chain leaves the run with no callable tools,
+OpenClaw stops before submitting the prompt to the model. This is intentional:
+an agent configured with a missing tool such as
+`agents.list[].tools.allow: ["query_db"]` should fail loudly until the plugin
+that registers `query_db` is enabled, not continue as a text-only agent.
+
 Tool policies support `group:*` shorthands that expand to multiple tools. See [Tool groups](/gateway/sandbox-vs-tool-policy-vs-elevated#tool-groups-shorthands) for the full list.
 
 Per-agent elevated overrides (`agents.list[].tools.elevated`) can further restrict elevated exec for specific agents. See [Elevated Mode](/tools/elevated) for details.
diff --git a/src/agents/pi-embedded-runner/run/attempt.ts b/src/agents/pi-embedded-runner/run/attempt.ts
index f3ed025fa00..b125e3d6235 100644
--- a/src/agents/pi-embedded-runner/run/attempt.ts
+++ b/src/agents/pi-embedded-runner/run/attempt.ts
@@ -108,6 +108,11 @@ import {
   toClientToolDefinitions,
 } from "../../pi-tool-definition-adapter.js";
 import { createOpenClawCodingTools, resolveToolLoopDetectionConfig } from "../../pi-tools.js";
+import {
+  resolveEffectiveToolPolicy,
+  resolveGroupToolPolicy,
+  resolveSubagentToolPolicyForSession,
+} from "../../pi-tools.policy.js";
 import { wrapStreamFnTextTransforms } from "../../plugin-text-transforms.js";
 import { describeProviderRequestRoutingSummary } from "../../provider-attribution.js";
 import { registerProviderStreamForModel } from "../../provider-stream.js";
@@ -126,10 +131,18 @@ import {
   applySkillEnvOverridesFromSnapshot,
   resolveSkillsPromptForRun,
 } from "../../skills.js";
+import {
+  isSubagentEnvelopeSession,
+  resolveSubagentCapabilityStore,
+} from "../../subagent-capabilities.js";
 import { resolveSystemPromptOverride } from "../../system-prompt-override.js";
 import { buildSystemPromptParams } from "../../system-prompt-params.js";
 import { buildSystemPromptReport } from "../../system-prompt-report.js";
 import { resolveAgentTimeoutMs } from "../../timeout.js";
+import {
+  buildEmptyExplicitToolAllowlistError,
+  collectExplicitToolAllowlistSources,
+} from "../../tool-allowlist-guard.js";
 import { UNKNOWN_TOOL_THRESHOLD } from "../../tool-loop-detection.js";
 import {
   resolveTranscriptPolicy,
@@ -451,6 +464,79 @@ export function applyEmbeddedAttemptToolsAllow<T extends { name: string }>(
   return tools.filter((tool) => allowSet.has(tool.name));
 }
 
+function collectAttemptExplicitToolAllowlistSources(params: {
+  config?: EmbeddedRunAttemptParams["config"];
+  sessionKey?: string;
+  sandboxSessionKey?: string;
+  agentId?: string;
+  modelProvider?: string;
+  modelId?: string;
+  messageProvider?: string;
+  agentAccountId?: string | null;
+  groupId?: string | null;
+  groupChannel?: string | null;
+  groupSpace?: string | null;
+  spawnedBy?: string | null;
+  senderId?: string | null;
+  senderName?: string | null;
+  senderUsername?: string | null;
+  senderE164?: string | null;
+  sandboxToolPolicy?: { allow?: string[]; deny?: string[] };
+  toolsAllow?: string[];
+}) {
+  const { agentId, globalPolicy, globalProviderPolicy, agentPolicy, agentProviderPolicy } =
+    resolveEffectiveToolPolicy({
+      config: params.config,
+      sessionKey: params.sessionKey,
+      agentId: params.agentId,
+      modelProvider: params.modelProvider,
+      modelId: params.modelId,
+    });
+  const groupPolicy = resolveGroupToolPolicy({
+    config: params.config,
+    sessionKey: params.sessionKey,
+    spawnedBy: params.spawnedBy,
+    messageProvider: params.messageProvider,
+    groupId: params.groupId,
+    groupChannel: params.groupChannel,
+    groupSpace: params.groupSpace,
+    accountId: params.agentAccountId,
+    senderId: params.senderId,
+    senderName: params.senderName,
+    senderUsername: params.senderUsername,
+    senderE164: params.senderE164,
+  });
+  const subagentStore = resolveSubagentCapabilityStore(params.sandboxSessionKey, {
+    cfg: params.config,
+  });
+  const subagentPolicy =
+    params.sandboxSessionKey &&
+    isSubagentEnvelopeSession(params.sandboxSessionKey, {
+      cfg: params.config,
+      store: subagentStore,
+    })
+      ? resolveSubagentToolPolicyForSession(params.config, params.sandboxSessionKey, {
+          store: subagentStore,
+        })
+      : undefined;
+  return collectExplicitToolAllowlistSources([
+    { label: "tools.allow", allow: globalPolicy?.allow },
+    { label: "tools.byProvider.allow", allow: globalProviderPolicy?.allow },
+    {
+      label: agentId ? `agents.${agentId}.tools.allow` : "agent tools.allow",
+      allow: agentPolicy?.allow,
+    },
+    {
+      label: agentId ? `agents.${agentId}.tools.byProvider.allow` : "agent tools.byProvider.allow",
+      allow: agentProviderPolicy?.allow,
+    },
+    { label: "group tools.allow", allow: groupPolicy?.allow },
+    { label: "sandbox tools.allow", allow: params.sandboxToolPolicy?.allow },
+    { label: "subagent tools.allow", allow: subagentPolicy?.allow },
+    { label: "runtime toolsAllow", allow: params.toolsAllow },
+  ]);
+}
+
 export async function runEmbeddedAttempt(
   params: EmbeddedRunAttemptParams,
 ): Promise<EmbeddedRunAttemptResult> {
@@ -802,6 +888,32 @@ export async function runEmbeddedAttempt(
       tools: effectiveTools,
       clientTools,
     });
+    const explicitToolAllowlistSources = collectAttemptExplicitToolAllowlistSources({
+      config: params.config,
+      sessionKey: params.sessionKey,
+      sandboxSessionKey,
+      agentId: sessionAgentId,
+      modelProvider: params.provider,
+      modelId: params.modelId,
+      messageProvider: params.messageChannel ?? params.messageProvider,
+      agentAccountId: params.agentAccountId,
+      groupId: params.groupId,
+      groupChannel: params.groupChannel,
+      groupSpace: params.groupSpace,
+      spawnedBy: params.spawnedBy,
+      senderId: params.senderId,
+      senderName: params.senderName,
+      senderUsername: params.senderUsername,
+      senderE164: params.senderE164,
+      sandboxToolPolicy: sandbox?.tools,
+      toolsAllow: params.toolsAllow,
+    });
+    const emptyExplicitToolAllowlistError = buildEmptyExplicitToolAllowlistError({
+      sources: explicitToolAllowlistSources,
+      callableToolNames: effectiveTools.map((tool) => tool.name),
+      toolsEnabled,
+      disableTools: params.disableTools,
+    });
     if (params.runtimePlan) {
       params.runtimePlan.tools.logDiagnostics(effectiveTools, runtimePlanModelContext);
     } else {
@@ -2031,6 +2143,12 @@ export async function runEmbeddedAttempt(
       let skipPromptSubmission = false;
       try {
         const promptStartedAt = Date.now();
+        if (emptyExplicitToolAllowlistError) {
+          promptError = emptyExplicitToolAllowlistError;
+          promptErrorSource = "precheck";
+          skipPromptSubmission = true;
+          log.warn(`[tools] ${emptyExplicitToolAllowlistError.message}`);
+        }
 
         // Run before_prompt_build hooks to allow plugins to inject prompt context.
         // Legacy compatibility: before_agent_start is also checked for context fields.
diff --git a/src/agents/tool-allowlist-guard.test.ts b/src/agents/tool-allowlist-guard.test.ts
new file mode 100644
index 00000000000..89e5f60e463
--- /dev/null
+++ b/src/agents/tool-allowlist-guard.test.ts
@@ -0,0 +1,75 @@
+import { describe, expect, it } from "vitest";
+import {
+  buildEmptyExplicitToolAllowlistError,
+  collectExplicitToolAllowlistSources,
+} from "./tool-allowlist-guard.js";
+
+describe("tool allowlist guard", () => {
+  it("fails closed when explicit allowlists resolve to no callable tools", () => {
+    const error = buildEmptyExplicitToolAllowlistError({
+      sources: [{ label: "tools.allow", entries: [" query_db "] }],
+      callableToolNames: [],
+      toolsEnabled: true,
+    });
+
+    expect(error?.message).toContain("No callable tools remain");
+    expect(error?.message).toContain("tools.allow: query_db");
+    expect(error?.message).toContain("no registered tools matched");
+  });
+
+  it("fails closed for runtime toolsAllow when tools are disabled", () => {
+    const error = buildEmptyExplicitToolAllowlistError({
+      sources: [{ label: "runtime toolsAllow", entries: ["query_db"] }],
+      callableToolNames: [],
+      toolsEnabled: true,
+      disableTools: true,
+    });
+
+    expect(error?.message).toContain("runtime toolsAllow: query_db");
+    expect(error?.message).toContain("tools are disabled for this run");
+  });
+
+  it("fails closed when the selected model cannot use requested tools", () => {
+    const error = buildEmptyExplicitToolAllowlistError({
+      sources: [{ label: "agents.db.tools.allow", entries: ["query_db"] }],
+      callableToolNames: [],
+      toolsEnabled: false,
+    });
+
+    expect(error?.message).toContain("agents.db.tools.allow: query_db");
+    expect(error?.message).toContain("the selected model does not support tools");
+  });
+
+  it("allows text-only runs without explicit allowlists", () => {
+    expect(
+      buildEmptyExplicitToolAllowlistError({
+        sources: [],
+        callableToolNames: [],
+        toolsEnabled: true,
+      }),
+    ).toBeNull();
+  });
+
+  it("allows explicit allowlists when at least one callable tool remains", () => {
+    expect(
+      buildEmptyExplicitToolAllowlistError({
+        sources: [{ label: "tools.allow", entries: ["read", "missing_tool"] }],
+        callableToolNames: ["read"],
+        toolsEnabled: true,
+      }),
+    ).toBeNull();
+  });
+
+  it("keeps source labels for config and runtime allowlists", () => {
+    const sources = collectExplicitToolAllowlistSources([
+      { label: "tools.allow", allow: [" read ", ""] },
+      { label: "runtime toolsAllow", allow: ["query_db"] },
+      { label: "tools.byProvider.allow" },
+    ]);
+
+    expect(sources).toEqual([
+      { label: "tools.allow", entries: ["read"] },
+      { label: "runtime toolsAllow", entries: ["query_db"] },
+    ]);
+  });
+});
diff --git a/src/agents/tool-allowlist-guard.ts b/src/agents/tool-allowlist-guard.ts
new file mode 100644
index 00000000000..33c356a4f43
--- /dev/null
+++ b/src/agents/tool-allowlist-guard.ts
@@ -0,0 +1,39 @@
+import { normalizeToolName } from "./tool-policy.js";
+
+export type ExplicitToolAllowlistSource = {
+  label: string;
+  entries: string[];
+};
+
+export function collectExplicitToolAllowlistSources(
+  sources: Array<{ label: string; allow?: string[] }>,
+): ExplicitToolAllowlistSource[] {
+  return sources.flatMap((source) => {
+    const entries = (source.allow ?? []).map((entry) => entry.trim()).filter(Boolean);
+    return entries.length ? [{ label: source.label, entries }] : [];
+  });
+}
+
+export function buildEmptyExplicitToolAllowlistError(params: {
+  sources: ExplicitToolAllowlistSource[];
+  callableToolNames: string[];
+  toolsEnabled: boolean;
+  disableTools?: boolean;
+}): Error | null {
+  const callableToolNames = params.callableToolNames.map(normalizeToolName).filter(Boolean);
+  if (params.sources.length === 0 || callableToolNames.length > 0) {
+    return null;
+  }
+  const requested = params.sources
+    .map((source) => `${source.label}: ${source.entries.map(normalizeToolName).join(", ")}`)
+    .join("; ");
+  const reason =
+    params.disableTools === true
+      ? "tools are disabled for this run"
+      : params.toolsEnabled
+        ? "no registered tools matched"
+        : "the selected model does not support tools";
+  return new Error(
+    `No callable tools remain after resolving explicit tool allowlist (${requested}); ${reason}. Fix the allowlist or enable the plugin that registers the requested tool.`,
+  );
+}