Add OpenAI-compatible llm.Provider adapter

internal/ai/llm/openai implements Provider over plain net/http against
POST {base}/chat/completions and GET {base}/models, with Bearer auth and
optional extra headers. Structured output uses response_format
json_schema (strict), json_object, or prompt-only with first-JSON-object
extraction, per the configured mode. Usage tokens, model and
finish_reason are parsed.

Errors map 401/403 to ErrAuth, 429 to ErrRateLimited with Retry-After
(seconds or HTTP-date), 408 and transport timeouts to ErrTimeout, and
other non-2xx statuses (and OpenRouter error bodies in a 2xx) to
ErrProvider with a key-redacted excerpt.

Contract tests replay OpenAI, OpenRouter and Ollama /v1 payloads from
httptest servers and assert the key and Authorization value never
appear in an error.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

internal/ai/llm/openai/testdata/ollama_chat_prompt_only.json

@@ -0,0 +1,18 @@
+{
+  "id": "chatcmpl-417",
+  "object": "chat.completion",
+  "created": 1741570322,
+  "model": "llama3.2",
+  "system_fingerprint": "fp_ollama",
+  "choices": [
+    {
+      "index": 0,
+      "message": {
+        "role": "assistant",
+        "content": "Sure! Here is a dinner {idea} for you:\n```json\n{\"dinners\": [{\"name\": \"Crock pot chili\", \"why\": \"A Quick dinner like your {Regular} stew.\"}]}\n```\nEnjoy {\"not\": \"this one\"}!"
+      },
+      "finish_reason": "stop"
+    }
+  ],
+  "usage": {"prompt_tokens": 512, "completion_tokens": 58, "total_tokens": 570}
+}

internal/ai/llm/openai/testdata/ollama_chat_text.json

@@ -0,0 +1,11 @@
+{
+  "id": "chatcmpl-419",
+  "object": "chat.completion",
+  "created": 1741570340,
+  "model": "llama3.2",
+  "system_fingerprint": "fp_ollama",
+  "choices": [
+    {"index": 0, "message": {"role": "assistant", "content": "Try a lemon orzo bake tonight."}, "finish_reason": "stop"}
+  ],
+  "usage": {"prompt_tokens": 40, "completion_tokens": 9, "total_tokens": 49}
+}

internal/ai/llm/openai/testdata/ollama_chat_truncated.json

@@ -0,0 +1,15 @@
+{
+  "id": "chatcmpl-418",
+  "object": "chat.completion",
+  "created": 1741570330,
+  "model": "llama3.2",
+  "system_fingerprint": "fp_ollama",
+  "choices": [
+    {
+      "index": 0,
+      "message": {"role": "assistant", "content": "{\"dinners\": [{\"name\": \"Crock pot ch"},
+      "finish_reason": "length"
+    }
+  ],
+  "usage": {"prompt_tokens": 512, "completion_tokens": 16, "total_tokens": 528}
+}

internal/ai/llm/openai/testdata/ollama_error_not_found.json

@@ -0,0 +1 @@
+{"error": {"message": "model \"llama9\" not found, try pulling it first", "type": "api_error", "param": null, "code": null}}

internal/ai/llm/openai/testdata/ollama_models.json

@@ -0,0 +1,7 @@
+{
+  "object": "list",
+  "data": [
+    {"id": "llama3.2:latest", "object": "model", "created": 1741000000, "owned_by": "library"},
+    {"id": "qwen2.5:7b", "object": "model", "created": 1740000000, "owned_by": "library"}
+  ]
+}

internal/ai/llm/openai/testdata/openai_chat_json_schema.json

@@ -0,0 +1,28 @@
+{
+  "id": "chatcmpl-B9MHDbslfkBeAs8l4bebGdFOJ6PeG",
+  "object": "chat.completion",
+  "created": 1741570283,
+  "model": "gpt-4o-2024-08-06",
+  "choices": [
+    {
+      "index": 0,
+      "message": {
+        "role": "assistant",
+        "content": "{\"dinners\":[{\"name\":\"Sheet-pan gnocchi\",\"why\":\"Like your Regular tray bake, but quicker.\"}]}",
+        "refusal": null,
+        "annotations": []
+      },
+      "logprobs": null,
+      "finish_reason": "stop"
+    }
+  ],
+  "usage": {
+    "prompt_tokens": 1117,
+    "completion_tokens": 46,
+    "total_tokens": 1163,
+    "prompt_tokens_details": {"cached_tokens": 0, "audio_tokens": 0},
+    "completion_tokens_details": {"reasoning_tokens": 0, "audio_tokens": 0, "accepted_prediction_tokens": 0, "rejected_prediction_tokens": 0}
+  },
+  "service_tier": "default",
+  "system_fingerprint": "fp_fc9f1d7035"
+}

internal/ai/llm/openai/testdata/openai_chat_refusal.json

@@ -0,0 +1,16 @@
+{
+  "id": "chatcmpl-B9MHDbslfkBeAs8l4bebGdFOJ6PeH",
+  "object": "chat.completion",
+  "created": 1741570290,
+  "model": "gpt-4o-2024-08-06",
+  "choices": [
+    {
+      "index": 0,
+      "message": {"role": "assistant", "content": null, "refusal": "I'm sorry, I can't help with that request.", "annotations": []},
+      "logprobs": null,
+      "finish_reason": "stop"
+    }
+  ],
+  "usage": {"prompt_tokens": 210, "completion_tokens": 11, "total_tokens": 221},
+  "system_fingerprint": "fp_fc9f1d7035"
+}

internal/ai/llm/openai/testdata/openai_error_invalid_key.json

@@ -0,0 +1,8 @@
+{
+  "error": {
+    "message": "Incorrect API key provided: {{KEY}}. You can find your API key at https://platform.openai.com/account/api-keys.",
+    "type": "invalid_request_error",
+    "param": null,
+    "code": "invalid_api_key"
+  }
+}

internal/ai/llm/openai/testdata/openai_error_server.json

@@ -0,0 +1 @@
+{"error": {"message": "The server had an error while processing your request. Authorization: Bearer {{KEY}}", "type": "server_error", "param": null, "code": null}}

internal/ai/llm/openai/testdata/openai_models.json

@@ -0,0 +1,7 @@
+{
+  "object": "list",
+  "data": [
+    {"id": "gpt-4o-2024-08-06", "object": "model", "created": 1722814719, "owned_by": "system"},
+    {"id": "gpt-4o-mini", "object": "model", "created": 1721172741, "owned_by": "system"}
+  ]
+}

internal/ai/llm/openai/testdata/openrouter_chat_json_object.json

@@ -0,0 +1,22 @@
+{
+  "id": "gen-1741570300-xQmT9vWb4kR2nP7sL3aE",
+  "provider": "OpenAI",
+  "model": "openai/gpt-4o-mini",
+  "object": "chat.completion",
+  "created": 1741570300,
+  "choices": [
+    {
+      "logprobs": null,
+      "finish_reason": "stop",
+      "native_finish_reason": "stop",
+      "index": 0,
+      "message": {
+        "role": "assistant",
+        "content": "{\n  \"dinners\": [{\"name\": \"Chicken tinga tacos\", \"why\": \"Close to your taco-night Regular.\"}]\n}",
+        "refusal": null,
+        "reasoning": null
+      }
+    }
+  ],
+  "usage": {"prompt_tokens": 842, "completion_tokens": 31, "total_tokens": 873}
+}

internal/ai/llm/openai/testdata/openrouter_chat_upstream_error.json

@@ -0,0 +1,5 @@
+{
+  "id": "gen-1741570311-aB3dE5fG7hJ9kL1mN2pQ",
+  "error": {"code": 502, "message": "Upstream error from provider", "metadata": {"provider_name": "Lambda", "raw": "upstream returned 502"}},
+  "user_id": "user_2abcDEF"
+}

internal/ai/llm/openai/testdata/openrouter_error_rate_limited.json

@@ -0,0 +1 @@
+{"error": {"code": 429, "message": "Rate limit exceeded: free-models-per-min. Key {{KEY}}", "metadata": {"headers": {"X-RateLimit-Limit": "20", "X-RateLimit-Remaining": "0"}}}}

internal/ai/llm/openai/testdata/openrouter_models.json

@@ -0,0 +1,22 @@
+{
+  "data": [
+    {
+      "id": "openai/gpt-4o-mini",
+      "canonical_slug": "openai/gpt-4o-mini",
+      "name": "OpenAI: GPT-4o-mini",
+      "created": 1721260800,
+      "description": "GPT-4o mini is OpenAI's newest model after GPT-4 Omni.",
+      "context_length": 128000,
+      "architecture": {"modality": "text+image->text", "input_modalities": ["text", "image"], "output_modalities": ["text"], "tokenizer": "GPT"},
+      "pricing": {"prompt": "0.00000015", "completion": "0.0000006"},
+      "supported_parameters": ["max_tokens", "temperature", "response_format", "structured_outputs"]
+    },
+    {
+      "id": "anthropic/claude-sonnet-5",
+      "name": "Anthropic: Claude Sonnet 5",
+      "created": 1767225600,
+      "context_length": 200000,
+      "supported_parameters": ["max_tokens", "temperature"]
+    }
+  ]
+}