fix: stop enforcing <final> for ollama (#16191) (thanks @Glucksberg)

CHANGELOG.md

@@ -8,6 +8,7 @@ Docs: https://docs.openclaw.ai
 
 - Security/Hooks: restrict hook transform modules to `~/.openclaw/hooks/transforms` (prevents path traversal/escape module loads via config). Config note: `hooks.transformsDir` must now be within that directory. Thanks @akhmittra.
 - Security/Hooks: ignore hook package manifest entries that point outside the package directory (prevents out-of-tree handler loads during hook discovery).
+- Ollama/Agents: avoid forcing `<final>` tag enforcement for Ollama models, which could suppress all output as `(no output)`. (#16191) Thanks @Glucksberg.
 
 ## 2026.2.13
 

src/utils/provider-utils.test.ts

@@ -2,7 +2,7 @@ import { describe, expect, it } from "vitest";
 import { isReasoningTagProvider } from "./provider-utils.js";
 
 describe("isReasoningTagProvider", () => {
-  it("returns false for ollama — native reasoning field, no tags needed (#2279)", () => {
+  it("returns false for ollama - native reasoning field, no tags needed (#2279)", () => {
     expect(isReasoningTagProvider("ollama")).toBe(false);
     expect(isReasoningTagProvider("Ollama")).toBe(false);
   });

src/utils/provider-utils.ts

@@ -14,7 +14,7 @@ export function isReasoningTagProvider(provider: string | undefined | null): boo
   const normalized = provider.trim().toLowerCase();
 
   // Check for exact matches or known prefixes/substrings for reasoning providers.
-  // Note: Ollama is intentionally excluded — its OpenAI-compatible endpoint
+  // Note: Ollama is intentionally excluded - its OpenAI-compatible endpoint
   // handles reasoning natively via the `reasoning` field in streaming chunks,
   // so tag-based enforcement is unnecessary and causes all output to be
   // discarded as "(no output)" (#2279).