ur-agent 1.24.1 → 1.25.1

package/docs/CODE_FEATURE_INVENTORY.md

@@ -977,7 +977,8 @@ The broader settings schema also covers:
 
 ## Model/provider behavior
 
-- This external build routes model calls through an Ollama provider shim.
+- This external build supports legal provider adapters for subscription CLIs,
+  explicit API-key providers, and local/OpenAI-compatible runtimes.
 - The Ollama router inspects local models and classifies likely coding/fast
   capabilities.
 - Auto-routing is on by default through `UR_OLLAMA_AUTO_ROUTE`.

package/docs/CONFIGURATION.md

@@ -2,9 +2,22 @@
 
 UR reads configuration from CLI flags, environment variables, and project or user settings files.
 
-## Model Provider
+## Model Providers
 
-UR runs models strictly through the local Ollama app. The default request endpoint is:
+UR-AGENT supports official provider access paths only:
+
+- Subscription/login CLI providers: Codex CLI, Claude Code CLI, Gemini CLI,
+  and Antigravity CLI where officially supported.
+- Explicit API providers: OpenAI, Anthropic, Gemini, OpenRouter, and
+  OpenAI-compatible endpoints.
+- Local providers: Ollama, LM Studio, llama.cpp, and vLLM OpenAI-compatible
+  server mode.
+
+UR-AGENT never scrapes browser sessions, extracts OAuth refresh tokens, reads
+hidden provider auth files, bypasses provider restrictions, or proxies consumer
+web sessions as APIs.
+
+The default local request endpoint is:
 
 ```text
 http://localhost:11434/api
@@ -12,6 +25,29 @@ http://localhost:11434/api
 
 Any model exposed by that Ollama app can be used, including local models and Ollama Cloud-backed models. UR does not call remote provider APIs directly and does not manage model API keys.
 
+Provider configuration commands:
+
+```sh
+ur provider list
+ur provider doctor
+ur provider status
+ur config set provider ollama
+ur config set provider openai-compatible
+ur config set provider.fallback ollama
+ur config set model <model>
+ur config set base_url <url>
+```
+
+API keys are not written to UR settings. Set them in the environment when you
+explicitly choose API mode:
+
+```sh
+OPENAI_API_KEY=...
+ANTHROPIC_API_KEY=...
+GEMINI_API_KEY=...
+OPENROUTER_API_KEY=...
+```
+
 ### Reconfiguring the Ollama host
 
 The endpoint can be changed from UR in three ways, in order of precedence:

package/docs/USAGE.md

@@ -45,7 +45,7 @@ ur -p \
   "summarize this project"
 ```
 
-## Models
+## Models And Providers
 
 The wrapper in `bin/ur.js` honors explicit model choices in this order:
 
@@ -82,7 +82,31 @@ Precedence: `--ollama-host` > `OLLAMA_HOST` env > `ollama.host` setting > `local
 `--discover-ollama` shows the picker every time but does **not** save the choice;
 use `ollama.host` in settings if you want plain `ur` to default to a LAN host.
 
-Models exposed by the chosen Ollama app are valid, including local models and Ollama Cloud-backed models. UR does not call provider APIs directly or manage model API keys.
+Models exposed by the chosen Ollama app are valid, including local models and
+Ollama Cloud-backed models.
+
+UR-AGENT also has explicit provider commands for legal access paths:
+
+```sh
+ur provider list
+ur provider status
+ur provider doctor
+ur auth chatgpt
+ur auth claude
+ur auth gemini
+ur auth antigravity
+ur config set provider ollama
+ur config set provider openai-compatible
+ur config set model <model>
+ur config set base_url <url>
+ur config set provider.fallback ollama
+```
+
+API modes are explicit and read keys only from environment variables:
+`OPENAI_API_KEY`, `ANTHROPIC_API_KEY`, `GEMINI_API_KEY`, and
+`OPENROUTER_API_KEY`. Subscription modes call official CLIs such as Codex,
+Claude Code, Gemini CLI, or Antigravity where supported. UR-AGENT never scrapes
+browser sessions, extracts OAuth tokens, or bypasses provider restrictions.
 
 ## Project Instructions
 
@@ -197,7 +221,7 @@ Interactive sessions include a compact bottom status bar when stdout is a real
 terminal:
 
 ```text
-UR-AGENT v1.24.0 | model: qwen3-coder:480b-cloud | mode: ask | branch: main | tasks: idle | Update: 1.23.3 -> 1.24.0 available
+UR-AGENT v1.25.1 | Provider: Ollama | Auth: local | model: qwen3-coder:480b-cloud | mode: ask | branch: main | tasks: idle | Update: 1.25.0 -> 1.25.1 available
 ```
 
 The bar is not rendered in non-interactive mode, CI, dumb terminals, or

package/docs/VALIDATION.md

@@ -18,7 +18,7 @@ You need:
 
 ```sh
 ur --version
-# expected: 1.24.1 (UR-AGENT)
+# expected: 1.25.1 (UR-AGENT)
 ```
 
 ## 0.1 Permission safety and context pack (1.19.0)

package/docs/providers.md

@@ -0,0 +1,104 @@
+# UR-AGENT providers
+
+UR-AGENT integrates official model access paths only. It supports subscription
+CLI login flows, explicit API-key providers, and local OpenAI-compatible
+runtimes. It does not implement hidden or unofficial authentication.
+
+## Legal auth policy
+
+UR-AGENT never:
+
+- scrapes browser cookies or browser sessions
+- extracts, copies, or reuses OAuth refresh tokens
+- reads hidden provider auth files directly
+- bypasses subscription, quota, region, product, or organization restrictions
+- proxies a consumer web session as an API
+- claims provider support unless the official CLI/API path works
+
+UR-AGENT stores only safe config: provider name, model name, base URL, command
+path, fallback preference, and non-secret preferences. API keys are read from
+environment variables only when the user explicitly selects API mode.
+
+## Provider matrix
+
+| Provider | Access type | Legal path |
+| --- | --- | --- |
+| ChatGPT/Codex | subscription | official Codex CLI login |
+| Claude Code | subscription | official Claude Code login |
+| Gemini CLI | subscription | official Gemini Code Assist login |
+| Antigravity | subscription | official Antigravity login, where supported |
+| OpenAI | API | `OPENAI_API_KEY` |
+| Anthropic Claude | API | `ANTHROPIC_API_KEY` |
+| Gemini | API | `GEMINI_API_KEY` |
+| OpenRouter | API/router | `OPENROUTER_API_KEY` |
+| Ollama | local | localhost Ollama runtime |
+| LM Studio | local | local OpenAI-compatible server |
+| llama.cpp | local | local OpenAI-compatible server |
+| vLLM | local/server | OpenAI-compatible server |
+
+## Commands
+
+```sh
+ur provider list
+ur provider status
+ur provider doctor
+ur provider doctor codex-cli
+ur auth chatgpt
+ur auth claude
+ur auth gemini
+ur auth antigravity
+ur config set provider codex-cli
+ur config set provider ollama
+ur config set provider openai-compatible
+ur config set model <model>
+ur config set base_url <url>
+ur config set provider.fallback ollama
+```
+
+`ur provider doctor` checks the selected provider. It reports installed/missing
+CLIs, official login status where available, API key presence for API providers,
+local endpoint reachability, detectable model availability, unsupported account
+type signals, and fallback configuration.
+
+Fallback is never silent by default. If the selected provider fails, UR-AGENT
+reports the selected provider, failure reason, suggested fix, and configured
+fallback option.
+
+## Subscription CLI providers
+
+- `codex-cli`: detects `codex --version`, uses `codex login` or
+  `codex login --device-auth`, and checks `codex login status`.
+- `claude-code-cli`: detects `claude --version`, uses `claude auth login`, and
+  checks `claude auth status` when available. If `ANTHROPIC_API_KEY` is set,
+  doctor warns that API-key mode may override subscription auth.
+- `gemini-cli`: launches only the official Gemini CLI flow. If the detected
+  path is an unsupported personal-account path, UR-AGENT prints a clear error.
+- `antigravity-cli`: launches only an installed official CLI command where
+  supported; UR-AGENT does not invent flags.
+
+## API and local providers
+
+API providers require explicit user selection and environment keys:
+
+```sh
+OPENAI_API_KEY=...
+ANTHROPIC_API_KEY=...
+GEMINI_API_KEY=...
+OPENROUTER_API_KEY=...
+```
+
+OpenAI-compatible endpoints can point at local or cloud endpoints:
+
+```sh
+ur config set provider openai-compatible
+ur config set base_url http://localhost:1234/v1
+ur config set model local-model-name
+```
+
+Local providers use their normal servers:
+
+- Ollama: `http://localhost:11434`
+- LM Studio: `http://localhost:1234/v1`
+- llama.cpp server mode: `http://localhost:8080/v1`
+- vLLM server mode: `http://localhost:8000/v1`
+

package/extensions/vscode-ur-inline-diffs/package.json

@@ -2,7 +2,7 @@
   "name": "ur-inline-diffs",
   "displayName": "UR Inline Diffs",
   "description": "Review UR inline diff bundles from .ur/ide/diffs inside VS Code.",
-  "version": "1.22.4",
+  "version": "1.25.1",
   "publisher": "ur-agent",
   "engines": {
     "vscode": "^1.92.0"

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ur-agent",
-  "version": "1.24.1",
+  "version": "1.25.1",
   "description": "UR-AGENT — autonomous engineering workflow engine (plan, execute, test, verify, document, benchmark, reproduce)",
   "type": "module",
   "packageManager": "bun@1.3.14",