fix(agents): remove unconditional "Tools are disabled" prompt injection in CLI runner

`runCliAgent()` unconditionally appended "Tools are disabled in this
session. Do not call tools." to `extraSystemPrompt` for every CLI
backend session. The intent was to prevent the LLM from calling
OpenClaw's embedded API tools (since CLI backends manage their own
tools natively). However, CLI agents like Claude Code interpret this
text as a blanket prohibition on ALL tools, including their own native
Bash, Read, and Write tools.

This caused silent failures across cron jobs, group chats, and DM
sessions when using any CLI backend: the agent would see the injected
text in the system prompt and refuse to execute tools, returning text
responses instead. Cron jobs reported `lastStatus: "ok"` despite the
agent failing to run scripts.

The fix removes the hardcoded string entirely. CLI backends already
receive `tools: []` (no OpenClaw embedded tools in the API call), so
the text was redundant at best.

Closes #44135

Co-Authored-By: Dhiman's Agentic Suite <dhiman.seal@hotmail.com>

src/agents/cli-runner.test.ts

@@ -231,6 +231,45 @@ describe("runCliAgent with process supervisor", () => {
     await loadFreshCliRunnerModuleForTest();
   });
 
+  it("does not inject hardcoded 'Tools are disabled' text into CLI arguments", async () => {
+    supervisorSpawnMock.mockResolvedValueOnce(
+      createManagedRun({
+        reason: "exit",
+        exitCode: 0,
+        exitSignal: null,
+        durationMs: 50,
+        stdout: "ok",
+        stderr: "",
+        timedOut: false,
+        noOutputTimedOut: false,
+      }),
+    );
+
+    await runCliAgent({
+      sessionId: "s1",
+      sessionFile: "/tmp/session.jsonl",
+      workspaceDir: "/tmp",
+      prompt: "Run: node script.mjs",
+      provider: "codex-cli",
+      model: "gpt-5.2-codex",
+      timeoutMs: 1_000,
+      runId: "run-no-tools-disabled",
+      extraSystemPrompt: "You are a helpful assistant.",
+    });
+
+    expect(supervisorSpawnMock).toHaveBeenCalledTimes(1);
+    const input = supervisorSpawnMock.mock.calls[0]?.[0] as {
+      argv?: string[];
+      input?: string;
+    };
+    // The CLI runner must not inject "Tools are disabled" into the system
+    // prompt passed to CLI backends. CLI backends (e.g., Claude Code CLI)
+    // manage their own native tools; the injected text caused them to
+    // refuse using their own tools. See: openclaw/openclaw#44135
+    const allArgs = (input.argv ?? []).join("\n");
+    expect(allArgs).not.toContain("Tools are disabled in this session");
+  });
+
   it("runs CLI through supervisor and returns payload", async () => {
     supervisorSpawnMock.mockResolvedValueOnce(
       createManagedRun({

src/agents/cli-runner.ts

@@ -128,12 +128,7 @@ export async function runCliAgent(params: {
   const normalizedModel = normalizeCliModel(modelId, backend);
   const modelDisplay = `${params.provider}/${modelId}`;
 
-  const extraSystemPrompt = [
-    params.extraSystemPrompt?.trim(),
-    "Tools are disabled in this session. Do not call tools.",
-  ]
-    .filter(Boolean)
-    .join("\n");
+  const extraSystemPrompt = params.extraSystemPrompt?.trim() ?? "";
 
   const sessionLabel = params.sessionKey ?? params.sessionId;
   const { bootstrapFiles, contextFiles } = await resolveBootstrapContextForRun({