@samooth/open-codex 0.2.7 → 0.2.11

package/README.md

@@ -111,7 +111,10 @@ development_ that understands and executes your repo.
 - **Zero setup** — bring your API key and it just works!
 - **Multiple AI providers** — use OpenAI, Gemini, OpenRouter, Ollama, xAI, DeepSeek, or Hugging Face!
 - **High Performance** — parallel tool execution and asynchronous file indexing for speed ✨
-- **Syntax Highlighting** — full terminal color support for code diffs and file contents 🎨
+- **Persistent Status Bar** — real-time visibility of model, provider, context usage, and session ID 📊
+- **Integrated Tool UI** — unified tool call and response boxes for a cleaner workflow 📦
+- **Syntax Highlighting** — full terminal color support for code diffs, search results, and file contents 🎨
+- **Interactive Search** — quickly filter through command history and past sessions 🔍
 - **Full auto-approval, while safe + secure** by running network-disabled and directory-sandboxed
 - **Multimodal** — pass in screenshots or diagrams to implement features ✨
 - **Dry Run mode** — preview all changes without actually modifying files or running commands!
@@ -290,55 +293,6 @@ npm link
 
 </details>
 
-### Alternative AI Providers
-
-This fork of Codex supports multiple AI providers:
-
-- `openai` (default)
-- `gemini`
-- `openrouter`
-- `ollama`
-- `xai`
-- `deepseek`
-- `hf` (Hugging Face)
-
-To use a different provider, set the `provider` key in your config file:
-
-```json
-{
-  "provider": "gemini"
-}
-```
-
-Or use the `--provider` flag, e.g. `codex --provider gemini`.
-
-#### Ollama Configuration
-
-When using the `ollama` provider, OpenCodex defaults to communicating with a local server at `http://localhost:11434`. You can customize this by setting the **`OLLAMA_BASE_URL`** environment variable:
-
-```bash
-export OLLAMA_BASE_URL="http://192.168.1.50:11434"
-```
-
-For embeddings (used in Project Memory and RAG), OpenCodex defaults to the `nomic-embed-text:latest` model for Ollama. Ensure you have pulled it:
-`ollama pull nomic-embed-text:latest`
-
-#### Dynamic Model Discovery
-
-For many providers, you can use the `/models` command within the interactive chat to see a list of available models and switch between them. For the **Hugging Face** provider, this dynamically fetches the latest `tool-use` compatible models directly from the Hugging Face Hub.
-
-Here’s a table of all providers and their default models:
-
-| Provider   | Environment Variable Required | Default Agentic Model | Default Full Context Model |
-|------------|-------------------------------|-----------------------|----------------------------|
-| openai     | `OPENAI_API_KEY`              | `o4-mini`             | `o3`                       |
-| gemini     | `GOOGLE_GENERATIVE_AI_API_KEY`| `gemini-3-pro-preview`| `gemini-2.5-pro`           |
-| openrouter | `OPENROUTER_API_KEY`          | `openai/o4-mini`      | `openai/o3`                |
-| ollama     | *Not required*                | *User must specify*   | *User must specify*        |
-| xai        | `XAI_API_KEY`                 | `grok-3-mini-beta`    | `grok-3-beta`              |
-| deepseek   | `DS_API_KEY`                  | `deepseek-chat`       | `deepseek-reasoner`        |
-| hf         | `HF_API_KEY`                  | `moonshotai/Kimi-K2.5`| `moonshotai/Kimi-K2.5`     |
-
 ---
 
 ## Configuration
@@ -352,9 +306,6 @@ Codex looks for config files in **`~/.codex/`** (either YAML or JSON format). Th
   "provider": "openai", // Default provider
   "approvalMode": "suggest", // or auto-edit, full-auto
   "fullAutoErrorMode": "ask-user", // or ignore-and-continue
-  "enableWebSearch": false, // default is false
-  "enableDeepThinking": false, // adds "Deep Thinking" prefix to prompt
-  "embeddingModel": "text-embedding-3-small", // Custom model for RAG/Memory
   "memory": {
     "enabled": true
   }
@@ -363,13 +314,57 @@ Codex looks for config files in **`~/.codex/`** (either YAML or JSON format). Th
 
 You can also define custom instructions:
 
-```markdown
+```md
 # ~/.codex/instructions.md
 
 - Always respond with emojis
 - Only use git commands if I explicitly mention you should
 ```
 
+### Alternative AI Providers
+
+This fork of Codex supports multiple AI providers:
+
+- openai (default)
+- gemini
+- openrouter
+- ollama
+- xai
+- deepseek
+- hf (Hugging Face)
+
+To use a different provider, set the `provider` key in your config file:
+
+```json
+{
+  "provider": "gemini"
+}
+```
+
+OR use the `--provider` flag. eg. `codex --provider gemini`
+
+#### Dynamic Model Discovery
+
+For many providers, you can use the `/models` command within the interactive chat to see a list of available models and switch between them. For the **Hugging Face** provider, this dynamically fetches the latest `tool-use` compatible models directly from the Hugging Face Hub.
+
+Here's a list of all the providers and their default models:
+
+| Provider   | Environment Variable Required | Default Agentic Model        | Default Full Context Model |
+| ---------- | ----------------------------- | ---------------------------- | -------------------------- |
+| openai     | OPENAI_API_KEY                | o4-mini                      | o3                         |
+| gemini     | GOOGLE_GENERATIVE_AI_API_KEY  | gemini-3-pro-preview         | gemini-2.5-pro             |
+| openrouter | OPENROUTER_API_KEY            | openai/o4-mini               | openai/o3                  |
+| ollama     | Not required                  | User must specify            | User must specify          |
+| xai        | XAI_API_KEY                   | grok-3-mini-beta             | grok-3-beta                |
+| deepseek   | DS_API_KEY                    | deepseek-chat                | deepseek-reasoner          |
+| hf         | HF_API_KEY                    | moonshotai/Kimi-K2.5         | moonshotai/Kimi-K2.5       |
+
+#### When using an alternative provider, make sure you have the correct environment variables set.
+
+```bash
+export GOOGLE_GENERATIVE_AI_API_KEY="your-gemini-api-key-here"
+```
+
 ---
 
 ## FAQ