You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@camel.apache.org by da...@apache.org on 2026/09/06 15:05:47 UTC
(camel) 01/01: CAMEL-24631: Ollama detection in camel doctor and TUI, LLM_BASE_URL support for OpenAI-compatible servers
This is an automated email from the ASF dual-hosted git repository.
davsclaus pushed a commit to branch feature/CAMEL-24631-ollama-doctor-tui
in repository https://gitbox.apache.org/repos/asf/camel.git
commit e0391d6570d4a15cfe6d0563e4c5fad3b6b66688
Author: Claus Ibsen <cl...@gmail.com>
AuthorDate: Sun Sep 6 17:02:56 2026 +0200
CAMEL-24631: Ollama detection in camel doctor and TUI, LLM_BASE_URL support for OpenAI-compatible servers
- camel doctor detects Ollama at localhost:11434 and lists pulled models
- TUI doctor popup shows Ollama status and model count; AI provider row
recognises local Ollama so it no longer warns when no API key is set
- F8 AI panel shows a rich setup guide (Ollama install for macOS+Linux,
memory table, OpenAI-compatible server section) when no provider is found
- LlmClient respects LLM_BASE_URL / OPENAI_BASE_URL so LLM_API_KEY can
point at any OpenAI-compatible server (LM Studio, vLLM, llama.cpp, …)
- Docs: camel-jbang-ai.adoc and camel-jbang-tui.adoc updated with Ollama
native install guide, model memory table (incl. hermes3), and
OpenAI-compatible server section
CAMEL-24631: remove small models from docs/guides and add DoctorPopup small-model warning
- Remove models <14B (llama3.2:3b, llama3.2:8b, hermes3:8b) from all
guides; only list ≥14B models suitable for tool calling
- Add prominent tool-calling warning to camel-jbang-tui.adoc and
camel-jbang-ai.adoc explaining models <14B don't reliably call tools
- DoctorPopup: show WARN icon and "F8 needs ≥14B" hint when all pulled
Ollama models are smaller than 14B parameters
- DoctorPopup: update no-models hint to suggest qwen2.5:14b
- AiPanel setup guide: update model table to ≥14B only with tool-calling
warning blockquote
Co-Authored-By: Claude Sonnet 4.6 <no...@anthropic.com>
---
.../modules/ROOT/pages/camel-jbang-ai.adoc | 68 +++++++++--
.../modules/ROOT/pages/camel-jbang-tui.adoc | 67 +++++++++++
.../camel/dsl/jbang/core/commands/Doctor.java | 58 ++++++++++
.../camel/dsl/jbang/core/commands/LlmClient.java | 8 +-
.../camel/dsl/jbang/core/commands/tui/AiPanel.java | 78 ++++++++++++-
.../dsl/jbang/core/commands/tui/DoctorPopup.java | 126 ++++++++++++++++++++-
6 files changed, 392 insertions(+), 13 deletions(-)
diff --git a/docs/user-manual/modules/ROOT/pages/camel-jbang-ai.adoc b/docs/user-manual/modules/ROOT/pages/camel-jbang-ai.adoc
index b68e7245cafe..baa4f1e33852 100644
--- a/docs/user-manual/modules/ROOT/pages/camel-jbang-ai.adoc
+++ b/docs/user-manual/modules/ROOT/pages/camel-jbang-ai.adoc
@@ -118,9 +118,10 @@ All commands auto-detect the LLM provider. The detection order is:
2. `CLOUD_ML_REGION` + `ANTHROPIC_VERTEX_PROJECT_ID` → Vertex AI (`ask` and `explain` only)
3. `AZURE_OPENAI_API_KEY` + `AZURE_OPENAI_ENDPOINT` → Azure OpenAI (uses the `api-key` header; optional `AZURE_OPENAI_DEPLOYMENT_NAME` and `AZURE_OPENAI_API_VERSION`)
4. `GEMINI_API_KEY` environment variable → Google Gemini native API (`generativelanguage.googleapis.com`). With `--api-type=gemini`, `GOOGLE_API_KEY` is also accepted.
-5. `OPENAI_API_KEY` or `LLM_API_KEY` → OpenAI-compatible API
-6. Ollama running via `camel infra` → local Ollama
-7. Ollama at `localhost:11434` → local Ollama
+5. `OPENAI_API_KEY` → OpenAI API (`api.openai.com`)
+6. `LLM_API_KEY` + optional `LLM_BASE_URL` (or `OPENAI_BASE_URL`) → any OpenAI-compatible API
+7. Ollama running via `camel infra` → local Ollama
+8. Ollama at `localhost:11434` → local Ollama
Override with explicit options:
@@ -134,16 +135,65 @@ camel ask "check health" --api-type=ollama --model=llama3.1
=== Using a local model with Ollama
-Start Ollama as a dev service and the CLI detects it automatically:
+Install Ollama natively for the best performance — the native binary uses GPU acceleration
+(Metal on macOS, CUDA/ROCm on Linux). Running Ollama through Docker (`camel infra run ollama`)
+bypasses the GPU and makes inference much slower.
[source,bash]
----
-camel infra run ollama
+# macOS
+brew install ollama
+
+# Linux
+curl -fsSL https://ollama.com/install.sh | sh
+
+# Pull a model and start asking
+ollama pull qwen2.5:32b
camel ask "what routes are running?"
----
-The default model is `llama3.2`. When using Ollama, the CLI checks what models are available
-locally and auto-selects a suitable one if a better model is installed.
+Ollama at `localhost:11434` is auto-detected — no environment variable needed.
+The CLI checks what models are available and auto-selects a suitable one.
+
+==== Model requirements
+
+`camel ask` and the TUI F8 panel rely on tool calling to inspect your running Camel process.
+Models smaller than ~14B do not reliably invoke tools and answer from training knowledge instead.
+Use at least a 14B model; 32B is recommended.
+
+[options="header"]
+|===
+| Model | RAM (Q4) | Notes
+| `qwen2.5:14b` | ~9 GB | Minimum recommended
+| `qwen2.5:32b` | ~20 GB | Best balance of speed and quality
+| `deepseek-r1:32b` | ~20 GB | Strong reasoning
+| `hermes3:70b` | ~43 GB | Excellent tool calling, needs 64 GB+
+| `llama3.3:70b` | ~43 GB | Best open model, needs 64 GB+
+|===
+
+On Apple Silicon, all RAM is unified — a 64 GB M-series Mac can run `llama3.3:70b` comfortably alongside the OS and other dev tools.
+
+=== Using an OpenAI-compatible local server
+
+Many local LLM servers expose an OpenAI-compatible API. Use `LLM_API_KEY` and `LLM_BASE_URL`
+to point the CLI at any of them:
+
+[source,bash]
+----
+export LLM_API_KEY=any-value # required but can be any non-empty string
+export LLM_BASE_URL=http://localhost:1234 # your server's base URL
+camel ask "what routes are running?"
+----
+
+`OPENAI_BASE_URL` is accepted as an alternative to `LLM_BASE_URL` (common in other tools).
+
+Common OpenAI-compatible servers:
-TIP: For best results with the `ask` command's tool-calling capabilities,
-use a model that supports function calling well (e.g., `llama3.1`, `qwen3`, `mistral-nemo`).
+[options="header"]
+|===
+| Server | Default port | Notes
+| https://lmstudio.ai[LM Studio] | 1234 | GUI app, Mac/Windows/Linux
+| https://github.com/vllm-project/vllm[vLLM] | 8000 | Production-grade, NVIDIA GPU
+| https://github.com/ggerganov/llama.cpp[llama.cpp server] | 8080 | Runs on CPU and GPU
+| https://gpt4all.io[GPT4All] | 4891 | Desktop app
+|===
diff --git a/docs/user-manual/modules/ROOT/pages/camel-jbang-tui.adoc b/docs/user-manual/modules/ROOT/pages/camel-jbang-tui.adoc
index eac4a6e96dc5..174c298f9810 100644
--- a/docs/user-manual/modules/ROOT/pages/camel-jbang-tui.adoc
+++ b/docs/user-manual/modules/ROOT/pages/camel-jbang-tui.adoc
@@ -738,6 +738,73 @@ server that lets AI coding assistants interact with the dashboard.
The TUI is fully functional on its own -- AI integration is entirely optional.
+=== Choosing an AI provider
+
+Press *F8* to open the built-in AI prompt panel. The panel auto-detects a provider in this order:
+
+1. `ANTHROPIC_API_KEY` → Anthropic Claude
+2. `CLOUD_ML_REGION` + `ANTHROPIC_VERTEX_PROJECT_ID` → Vertex AI
+3. `AZURE_OPENAI_API_KEY` + `AZURE_OPENAI_ENDPOINT` → Azure OpenAI
+4. `GEMINI_API_KEY` → Google Gemini
+5. `OPENAI_API_KEY` → OpenAI
+6. `LLM_API_KEY` + optional `LLM_BASE_URL` → any OpenAI-compatible server
+7. Ollama at `localhost:11434` → local Ollama (auto-detected, no key needed)
+
+Press *Ctrl+P* inside the AI panel to switch provider or model at any time.
+
+==== Using Ollama (local, no API key)
+
+Install Ollama natively for best performance — the native binary uses GPU acceleration
+(Metal on macOS, CUDA/ROCm on Linux):
+
+[source,bash]
+----
+# macOS
+brew install ollama
+
+# Linux
+curl -fsSL https://ollama.com/install.sh | sh
+
+# Pull a model — then open the TUI and press F8
+ollama pull qwen2.5:32b
+camel tui
+----
+
+Ollama at `localhost:11434` is auto-detected. No configuration needed.
+
+IMPORTANT: The F8 AI panel works by invoking built-in tools to inspect your running Camel process.
+Models smaller than ~14B do not reliably call tools and answer from training knowledge instead.
+Use at least a 14B model; 32B is recommended.
+
+*Models that work well* (tool-calling capable, ≥14B, default Q4_K_M quantization):
+
+[options="header"]
+|===
+| Model | RAM | Notes
+| `qwen2.5:14b` | ~9 GB | Minimum recommended
+| `qwen2.5:32b` | ~20 GB | Best balance of speed and quality
+| `deepseek-r1:32b` | ~20 GB | Strong reasoning
+| `hermes3:70b` | ~43 GB | Excellent tool calling, needs 64 GB+
+| `llama3.3:70b` | ~43 GB | Best open model, needs 64 GB+
+|===
+
+NOTE: `camel infra run ollama` runs Ollama in Docker and bypasses GPU acceleration,
+making inference significantly slower. Native install is preferred for development.
+
+==== Using an OpenAI-compatible local server
+
+Set `LLM_API_KEY` and `LLM_BASE_URL` to connect to any OpenAI-compatible server
+(LM Studio, vLLM, llama.cpp, GPT4All, …):
+
+[source,bash]
+----
+export LLM_API_KEY=any-value
+export LLM_BASE_URL=http://localhost:1234
+camel tui
+----
+
+`OPENAI_BASE_URL` is also accepted as an alternative to `LLM_BASE_URL`.
+
=== Why This Matters
When an AI agent connects to the TUI via MCP, it gains the same level of visibility that you
diff --git a/dsl/camel-jbang/camel-jbang-core/src/main/java/org/apache/camel/dsl/jbang/core/commands/Doctor.java b/dsl/camel-jbang/camel-jbang-core/src/main/java/org/apache/camel/dsl/jbang/core/commands/Doctor.java
index 98b25fd985de..a72a6981a78b 100644
--- a/dsl/camel-jbang/camel-jbang-core/src/main/java/org/apache/camel/dsl/jbang/core/commands/Doctor.java
+++ b/dsl/camel-jbang/camel-jbang-core/src/main/java/org/apache/camel/dsl/jbang/core/commands/Doctor.java
@@ -19,7 +19,13 @@ package org.apache.camel.dsl.jbang.core.commands;
import java.io.File;
import java.io.OutputStream;
import java.net.ServerSocket;
+import java.net.URI;
+import java.net.http.HttpClient;
+import java.net.http.HttpRequest;
+import java.net.http.HttpResponse;
import java.nio.file.Path;
+import java.time.Duration;
+import java.util.ArrayList;
import java.util.List;
import java.util.Optional;
import java.util.Set;
@@ -31,6 +37,9 @@ import org.apache.camel.dsl.jbang.core.common.VersionHelper;
import org.apache.camel.tooling.maven.MavenDownloaderImpl;
import org.apache.camel.tooling.maven.MavenResolutionException;
import org.apache.camel.util.StringHelper;
+import org.apache.camel.util.json.JsonArray;
+import org.apache.camel.util.json.JsonObject;
+import org.apache.camel.util.json.Jsoner;
import picocli.CommandLine.Command;
@Command(name = "doctor", description = "Checks the environment and reports potential issues",
@@ -55,6 +64,7 @@ public class Doctor extends CamelCommand {
checkJBang();
checkMavenRepository();
checkContainerRuntime();
+ checkOllama();
checkCommonPorts();
checkDiskSpace();
@@ -133,6 +143,54 @@ public class Doctor extends CamelCommand {
printer().printf(" Container: Not found (optional — needed for running external infra services)%n");
}
+ private void checkOllama() {
+ try {
+ HttpClient client = HttpClient.newBuilder()
+ .connectTimeout(Duration.ofSeconds(3))
+ .build();
+ HttpRequest request = HttpRequest.newBuilder()
+ .uri(URI.create("http://localhost:11434/api/tags"))
+ .timeout(Duration.ofSeconds(3))
+ .GET()
+ .build();
+ HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
+ if (response.statusCode() == 200) {
+ List<String> models = parseOllamaModels(response.body());
+ if (models.isEmpty()) {
+ printer().printf(" Ollama: Running at localhost:11434 — no models pulled yet%n");
+ } else {
+ printer().printf(" Ollama: Running at localhost:11434 — models: %s%n",
+ String.join(", ", models));
+ }
+ } else {
+ printer().printf(" Ollama: Not detected (optional — start for local AI with F8 in TUI)%n");
+ }
+ } catch (Exception e) {
+ printer().printf(" Ollama: Not detected (optional — start for local AI with F8 in TUI)%n");
+ }
+ }
+
+ private List<String> parseOllamaModels(String json) {
+ List<String> names = new ArrayList<>();
+ try {
+ JsonObject root = (JsonObject) Jsoner.deserialize(json);
+ JsonArray models = (JsonArray) root.get("models");
+ if (models != null) {
+ for (Object entry : models) {
+ if (entry instanceof JsonObject model) {
+ Object name = model.get("name");
+ if (name != null) {
+ names.add(name.toString());
+ }
+ }
+ }
+ }
+ } catch (Exception e) {
+ // ignore parse errors — not critical
+ }
+ return names;
+ }
+
private void checkCommonPorts() {
StringBuilder conflicts = new StringBuilder();
for (int port : new int[] { 8080, 8443, 9090 }) {
diff --git a/dsl/camel-jbang/camel-jbang-core/src/main/java/org/apache/camel/dsl/jbang/core/commands/LlmClient.java b/dsl/camel-jbang/camel-jbang-core/src/main/java/org/apache/camel/dsl/jbang/core/commands/LlmClient.java
index 94f94b788887..8ebc1b78fcb1 100644
--- a/dsl/camel-jbang/camel-jbang-core/src/main/java/org/apache/camel/dsl/jbang/core/commands/LlmClient.java
+++ b/dsl/camel-jbang/camel-jbang-core/src/main/java/org/apache/camel/dsl/jbang/core/commands/LlmClient.java
@@ -1795,7 +1795,13 @@ public class LlmClient {
apiKey = key;
openAiAuthMode = OpenAiAuthMode.bearer;
if (url == null || url.isBlank()) {
- url = "https://api.openai.com";
+ // LLM_BASE_URL / OPENAI_BASE_URL let users point at any OpenAI-compatible
+ // server (LM Studio, vLLM, LocalAI, Jan, …) without a CLI flag
+ String baseUrl = System.getenv("LLM_BASE_URL");
+ if (baseUrl == null || baseUrl.isBlank()) {
+ baseUrl = System.getenv("OPENAI_BASE_URL");
+ }
+ url = (baseUrl != null && !baseUrl.isBlank()) ? stripTrailingSlash(baseUrl) : "https://api.openai.com";
}
return true;
}
diff --git a/dsl/camel-jbang/camel-jbang-plugin-tui/src/main/java/org/apache/camel/dsl/jbang/core/commands/tui/AiPanel.java b/dsl/camel-jbang/camel-jbang-plugin-tui/src/main/java/org/apache/camel/dsl/jbang/core/commands/tui/AiPanel.java
index fda8cd5ff4dd..904455a6b085 100644
--- a/dsl/camel-jbang/camel-jbang-plugin-tui/src/main/java/org/apache/camel/dsl/jbang/core/commands/tui/AiPanel.java
+++ b/dsl/camel-jbang/camel-jbang-plugin-tui/src/main/java/org/apache/camel/dsl/jbang/core/commands/tui/AiPanel.java
@@ -1054,7 +1054,11 @@ class AiPanel {
StringBuilder md = new StringBuilder();
if (initError != null) {
- md.append("**Error:** ").append(initError).append("\n\n");
+ if (initError.startsWith("No LLM service reachable")) {
+ md.append(buildAiSetupGuide());
+ } else {
+ md.append("**Error:** ").append(initError).append("\n\n");
+ }
} else if (conversation.isEmpty() && !thinking.get() && !slashHintsVisible) {
frame.renderWidget(
Paragraph.from(Line.from(Span.styled("Ask a question about your Camel application...", Style.EMPTY.dim()))),
@@ -1556,6 +1560,78 @@ class AiPanel {
frame.renderWidget(Paragraph.from(new dev.tamboui.text.Text(lines, dev.tamboui.layout.Alignment.LEFT)), area);
}
+ private String buildAiSetupGuide() {
+ return """
+ ## AI Assistant — Getting Started
+
+ No LLM provider was detected. Choose one of the options below, then press **F8** to reopen this panel.
+
+ > **Tool calling is required.** This panel inspects your Camel process by
+ > invoking built-in tools. Models smaller than ~14B do not reliably call
+ > tools and will answer from training knowledge instead — use at least 14B,
+ > 32B recommended.
+
+ ---
+
+ ### Option A: Local — Ollama (no API key needed)
+
+ Run models entirely on your machine — no data leaves your host.
+
+ ```
+ # macOS
+ brew install ollama
+
+ # Linux
+ curl -fsSL https://ollama.com/install.sh | sh
+
+ # then on both:
+ ollama serve # start the daemon (skip if auto-started)
+ ollama pull qwen2.5:32b # recommended
+ ```
+
+ Ollama is auto-detected at `localhost:11434` — no configuration needed.
+
+ **Models that work well** (tool-calling capable, ≥14B):
+
+ | Model | RAM | Notes |
+ |---|---|---|
+ | qwen2.5:14b | ~9 GB | Minimum recommended |
+ | qwen2.5:32b | ~20 GB | Best balance of speed and quality |
+ | deepseek-r1:32b | ~20 GB | Strong reasoning |
+ | hermes3:70b | ~43 GB | Excellent tool calling, needs 64 GB+ |
+ | llama3.3:70b | ~43 GB | Best open model, needs 64 GB+ |
+
+ **Tip:** Install Ollama natively — `camel infra run ollama` uses Docker and
+ loses GPU acceleration (Metal on macOS, CUDA on Linux), making inference
+ much slower. Native install is always preferred for development use.
+
+ ---
+
+ ### Option B: Cloud provider (API key required)
+
+ Set one environment variable before starting the TUI:
+
+ | Variable | Provider |
+ |---|---|
+ | `ANTHROPIC_API_KEY` | Claude |
+ | `OPENAI_API_KEY` | OpenAI (GPT-4o etc.) |
+ | `GEMINI_API_KEY` | Gemini |
+ | `AZURE_OPENAI_API_KEY` + `AZURE_OPENAI_ENDPOINT` | Azure OpenAI |
+ | `WATSONX_APIKEY` | IBM watsonx.ai |
+
+ For any **OpenAI-compatible** server (LM Studio, vLLM, llama.cpp, GPT4All, …):
+
+ ```
+ export LLM_API_KEY=any-value
+ export LLM_BASE_URL=http://localhost:1234
+ ```
+
+ `OPENAI_BASE_URL` is also supported as an alternative to `LLM_BASE_URL`.
+
+ Or press **Ctrl+P** to select and configure a provider now.
+ """;
+ }
+
private String buildSystemPrompt() {
StringBuilder sb = new StringBuilder();
sb.append("You are an Apache Camel assistant running inside the Camel TUI terminal console. ");
diff --git a/dsl/camel-jbang/camel-jbang-plugin-tui/src/main/java/org/apache/camel/dsl/jbang/core/commands/tui/DoctorPopup.java b/dsl/camel-jbang/camel-jbang-plugin-tui/src/main/java/org/apache/camel/dsl/jbang/core/commands/tui/DoctorPopup.java
index 9623f115eac6..0c3351790e56 100644
--- a/dsl/camel-jbang/camel-jbang-plugin-tui/src/main/java/org/apache/camel/dsl/jbang/core/commands/tui/DoctorPopup.java
+++ b/dsl/camel-jbang/camel-jbang-plugin-tui/src/main/java/org/apache/camel/dsl/jbang/core/commands/tui/DoctorPopup.java
@@ -19,6 +19,11 @@ package org.apache.camel.dsl.jbang.core.commands.tui;
import java.io.File;
import java.io.OutputStream;
import java.net.ServerSocket;
+import java.net.URI;
+import java.net.http.HttpClient;
+import java.net.http.HttpRequest;
+import java.net.http.HttpResponse;
+import java.time.Duration;
import java.util.ArrayList;
import java.util.List;
import java.util.Set;
@@ -42,6 +47,9 @@ import org.apache.camel.catalog.DefaultCamelCatalog;
import org.apache.camel.dsl.jbang.core.common.VersionHelper;
import org.apache.camel.tooling.maven.MavenDownloaderImpl;
import org.apache.camel.tooling.maven.MavenResolutionException;
+import org.apache.camel.util.json.JsonArray;
+import org.apache.camel.util.json.JsonObject;
+import org.apache.camel.util.json.Jsoner;
import static org.apache.camel.dsl.jbang.core.commands.tui.TuiHelper.hintLast;
@@ -70,6 +78,7 @@ class DoctorPopup {
checkJBang(lines);
checkMavenRepository(lines);
checkContainerRuntime(lines);
+ checkOllama(lines);
checkCommonPorts(lines);
checkDiskSpace(lines);
checkAiProvider(lines);
@@ -277,6 +286,117 @@ class DoctorPopup {
}
}
+ private void checkOllama(List<Line> result) {
+ try {
+ HttpClient client = HttpClient.newBuilder()
+ .connectTimeout(Duration.ofSeconds(3))
+ .build();
+ HttpRequest request = HttpRequest.newBuilder()
+ .uri(URI.create("http://localhost:11434/api/tags"))
+ .timeout(Duration.ofSeconds(3))
+ .GET()
+ .build();
+ HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());
+ if (response.statusCode() == 200) {
+ List<String> models = parseOllamaModels(response.body());
+ if (models.isEmpty()) {
+ result.add(Line.from(
+ Span.raw(TuiIcons.indent(TuiIcons.MCP)),
+ Span.styled(String.format("%-14s", "Ollama"), Theme.muted()),
+ Span.raw(String.format("%-30s", "Running — no models pulled yet")),
+ Span.raw(" " + TuiIcons.WARN)));
+ result.add(Line.from(Span.styled(" Run: ollama pull qwen2.5:14b",
+ Style.EMPTY.dim())));
+ } else {
+ boolean allSmall = models.stream().allMatch(DoctorPopup::isSmallModel);
+ String icon = allSmall ? TuiIcons.WARN : TuiIcons.OK;
+ result.add(Line.from(
+ Span.raw(TuiIcons.indent(TuiIcons.MCP)),
+ Span.styled(String.format("%-14s", "Ollama"), Theme.muted()),
+ Span.raw(String.format("%-30s", models.size() + " model(s) available")),
+ Span.raw(" " + icon)));
+ result.add(Line.from(Span.styled(
+ " " + TuiHelper.truncate(String.join(", ", models), 44),
+ Style.EMPTY.dim())));
+ if (allSmall) {
+ result.add(Line.from(Span.styled(
+ " F8 needs ≥14B — run: ollama pull qwen2.5:14b",
+ Style.EMPTY.dim())));
+ }
+ }
+ } else {
+ result.add(Line.from(
+ Span.raw(TuiIcons.indent(TuiIcons.MCP)),
+ Span.styled(String.format("%-14s", "Ollama"), Theme.muted()),
+ Span.raw(String.format("%-30s", "Not running (optional)")),
+ Span.raw(" " + TuiIcons.WARN)));
+ result.add(Line.from(Span.styled(" Run: ollama serve",
+ Style.EMPTY.dim())));
+ }
+ } catch (Exception e) {
+ result.add(Line.from(
+ Span.raw(TuiIcons.indent(TuiIcons.MCP)),
+ Span.styled(String.format("%-14s", "Ollama"), Theme.muted()),
+ Span.raw(String.format("%-30s", "Not running (optional)")),
+ Span.raw(" " + TuiIcons.WARN)));
+ result.add(Line.from(Span.styled(" Run: ollama serve",
+ Style.EMPTY.dim())));
+ }
+ }
+
+ private boolean isOllamaRunning() {
+ try {
+ HttpClient client = HttpClient.newBuilder()
+ .connectTimeout(Duration.ofSeconds(2))
+ .build();
+ HttpRequest request = HttpRequest.newBuilder()
+ .uri(URI.create("http://localhost:11434/api/tags"))
+ .timeout(Duration.ofSeconds(2))
+ .GET()
+ .build();
+ return client.send(request, HttpResponse.BodyHandlers.discarding()).statusCode() == 200;
+ } catch (Exception e) {
+ return false;
+ }
+ }
+
+ private static boolean isSmallModel(String name) {
+ // heuristic: extract the size suffix like :3b, :7b, :8b, :11b from the model name
+ int colon = name.lastIndexOf(':');
+ String tag = colon >= 0 ? name.substring(colon + 1).toLowerCase() : "";
+ if (tag.matches("\\d+b.*")) {
+ int b = tag.indexOf('b');
+ try {
+ int params = Integer.parseInt(tag.substring(0, b));
+ return params < 14;
+ } catch (NumberFormatException e) {
+ return false;
+ }
+ }
+ return false;
+ }
+
+ private List<String> parseOllamaModels(String json) {
+ List<String> names = new ArrayList<>();
+ try {
+ JsonObject root = (JsonObject) Jsoner.deserialize(json);
+ JsonArray models = (JsonArray) root.get("models");
+ if (models != null) {
+ for (Object entry : models) {
+ if (entry instanceof JsonObject model) {
+ Object name = model.get("name");
+ if (name != null) {
+ names.add(name.toString());
+ }
+ }
+ }
+ }
+ } catch (Exception e) {
+ // ignore parse errors
+ }
+ return names;
+ }
+
private static boolean envSet(String name) {
String v = System.getenv(name);
return v != null && !v.isBlank();
@@ -298,6 +418,8 @@ class DoctorPopup {
provider = "watsonx.ai";
} else if (envSet("LLM_API_KEY")) {
provider = "Custom (LLM_API_KEY)";
+ } else if (isOllamaRunning()) {
+ provider = "Ollama (local)";
}
if (provider != null) {
result.add(Line.from(
@@ -309,10 +431,10 @@ class DoctorPopup {
result.add(Line.from(
Span.raw(TuiIcons.indent(TuiIcons.MCP)),
Span.styled(String.format("%-14s", "AI"), Theme.muted()),
- Span.raw(String.format("%-30s", "No API key configured")),
+ Span.raw(String.format("%-30s", "No AI provider configured")),
Span.raw(" " + TuiIcons.WARN)));
result.add(Line.from(Span.styled(
- " Set ANTHROPIC_API_KEY, AZURE_OPENAI_*, GEMINI_API_KEY, OPENAI_API_KEY, or WATSONX_APIKEY",
+ " Set ANTHROPIC_API_KEY, OPENAI_API_KEY, GEMINI_API_KEY, AZURE_OPENAI_*, WATSONX_APIKEY, LLM_API_KEY+LLM_BASE_URL, or start Ollama",
Style.EMPTY.dim())));
}
}