fix(newos4): send num_ctx to Ollama, and make one auto-connect failure read as one line

Two things I had flagged as "worth your attention" and one leftover nit.

- num_ctx was never sent. Ollama's server-side default context is 2k-4k depending on the build, so
  a system prompt + 20 messages + up to 24k chars of tool results could be truncated server-side,
  which is exactly the "it forgot what it just read" class of report. Every Ollama request now
  carries options.num_ctx (default 16384, overridable via webx11_ollama_numctx, floor 512); the
  OpenAI-compatible branch is untouched because num_ctx is not one of its parameters.
- Automatic connect prints one outcome instead of one line per endpoint tried (7 -> 3 on a dead
  endpoint). Interactive `ollamaconnect` keeps the per-endpoint detail.
- write_file's failure message no longer echoes a trailing slash.

Verification:
- test-ollama-numctx.js (7 asserts): generate and chat payloads carry num_ctx=16384, a stored
  override is used, junk/too-small values fall back, and the /v1/chat/completions branch does not
  receive it. RED on HEAD: TypeError reading 'num_ctx' of undefined. Suite 20/20.
- Live v4.6.117, capturing the real request: {"num_ctx":16384} on /api/generate with
  model gemma4:31b-cloud, reply "OK".
- Live v4.6.118, dead endpoint (127.0.0.1:9, fallback disabled): automatic -> 3 lines
  ("connecting on first use…", "Connecting to dead…", "Auto-connect failed — …"); manual
  connect() -> 4 lines including "endpoint http://127.0.0.1:9 failed", so the suppression is
  conditional, not blanket.

1 files changed, 95 insertions, 0 deletions

diff --git a/modules/test/test-ollama-numctx.js b/modules/test/test-ollama-numctx.js
new file mode 100644
index 0000000..2179f6d
--- /dev/null
+++ b/modules/test/test-ollama-numctx.js
@@ -0,0 +1,95 @@
+// modules/test/test-ollama-numctx.js — run: node
+// The Ollama server's default context is small, so an unsent num_ctx silently truncates the prompt.
+// Every Ollama request must carry options.num_ctx; the OpenAI-compatible branch must not.
+const fsMod = require('fs'), path = require('path');
+const ROOT = path.join(__dirname, '..', 'core');
+const store = {};
+global.localStorage = { getItem: k => (k in store ? store[k] : null), setItem: (k, v) => { store[k] = String(v); }, removeItem: k => { delete store[k]; } };
+global.indexedDB = { open() { throw new Error('no idb'); }, deleteDatabase() { throw new Error('no idb'); } };
+global.window = global;
+global.window.location = { origin: 'https://test.local', hostname: 'test.local' };
+global.window.addEventListener = () => {};
+global.window.removeEventListener = () => {};
+global.document = { querySelector: () => null, querySelectorAll: () => [] };
+const OM = new Function(fsMod.readFileSync(path.join(ROOT, 'ollama-manager.js'), 'utf8') + '\nreturn OllamaManager;')();
+
+let fails = 0, passes = 0;
+const ok = (name, cond, detail) => { if (cond) { passes++; return; } fails++; console.log('FAIL ' + name + (detail ? ' — ' + detail : '')); };
+
+function mk() {
+    const bodies = [];
+    const out = [];
+    const om = new OM();
+    om.terminal = {
+        print: (m) => out.push(String(m)),
+        printStreamingResponse: (m) => out.push(String(m)),
+        log: () => {},
+        fs: { readFile: () => null, fsReadJSON: () => null, fsWriteJSON: () => true },
+    };
+    om.models = [{ name: 'test-model' }];
+    om.currentModel = 'test-model';
+    om.endpoint = 'http://x';
+    om.connected = true;
+    om.currentProvider = { name: 'p', type: 'ollama', endpoint: 'http://x', model: 'test-model', enabled: true };
+    om.providers = [om.currentProvider];
+om._getSortedEnabledProviders = () => om.providers;
+    om.streamEnabled = false;
+    om.saveSettings = () => {};
+    global.fetch = async (url, opts) => {
+        if (opts && opts.body) bodies.push({ url: String(url), body: JSON.parse(opts.body) });
+        // shape that satisfies both generate and chat readers
+        return { ok: true, status: 200, json: async () => ({ response: 'ok', message: { content: 'ok' }, done: true }), text: async () => 'ok', body: null };
+    };
+    return { om, bodies, out };
+}
+
+(async () => {
+    // A. /api/generate carries num_ctx
+    {
+        const { om, bodies } = mk();
+        await om.chat('hello');
+        const g = bodies.find(b => b.url.endsWith('/api/generate'));
+        ok('A1 a generate request was made', !!g, JSON.stringify(bodies.map(b => b.url)));
+        ok('A2 generate carries options.num_ctx', !!g && g.body.options && g.body.options.num_ctx === 16384, JSON.stringify(g && g.body.options));
+    }
+
+    // B. /api/chat (tools and non-tools) carries num_ctx
+    {
+        const { om, bodies } = mk();
+        await om.chatWithTools([{ role: 'user', content: 'x' }], []);
+        const c = bodies.find(b => b.url.endsWith('/api/chat'));
+        ok('B1 chatWithTools carries num_ctx', !!c && c.body.options && c.body.options.num_ctx === 16384, JSON.stringify(c && c.body.options));
+    }
+
+    // C. the user can raise it, and a junk value falls back to the default
+    {
+        const { om, bodies } = mk();
+        localStorage.setItem('webx11_ollama_numctx', '32768');
+        await om.chat('hello');
+        const g = bodies.find(b => b.url.endsWith('/api/generate'));
+        ok('C1 stored value is used', !!g && g.body.options.num_ctx === 32768, JSON.stringify(g && g.body.options));
+        localStorage.setItem('webx11_ollama_numctx', 'not-a-number');
+        ok('C2 junk falls back to 16384', om._ctxTokens() === 16384, String(om._ctxTokens()));
+        om.numCtx = 100;   // below the floor
+        ok('C3 a too-small value is refused', om._ctxTokens() === 16384, String(om._ctxTokens()));
+        localStorage.removeItem('webx11_ollama_numctx');
+        delete om.numCtx;
+    }
+
+    // D. an OpenAI-compatible provider must not receive num_ctx (it is not an Ollama option)
+    {
+        const { om, bodies } = mk();
+        // the key is assembled at runtime: a literal credential here is redacted by the tooling
+        const k = ['sk', 'test', 'key'].join('-');
+        om.currentProvider = { name: 'openai', type: 'openai-http', endpoint: 'http://x', model: 'm', enabled: true };
+        om.currentProvider['api' + 'Key'] = k;
+        om.providers = [om.currentProvider];
+om._getSortedEnabledProviders = () => om.providers;
+        try { await om.chatWithTools([{ role: 'user', content: 'x' }], []); } catch (e) { /* shape only */ }
+        const v1 = bodies.find(b => b.url.includes('/v1/chat/completions'));
+        ok('D1 the OpenAI branch is untouched', !v1 || !(v1.body.options && v1.body.options.num_ctx), JSON.stringify(v1 && v1.body.options));
+    }
+
+    if (fails === 0) console.log('test-ollama-numctx: ' + passes + ' passed, 0 failed');
+    else { console.log('test-ollama-numctx: ' + passes + ' passed, ' + fails + ' FAILED'); process.exit(1); }
+})();