logging: migrate LSP logs to global singleton (internal/logging); use consistent colors/prefix; refactor LLM provider to use global logger and remove per-client loggers

1 files changed, 16 insertions, 16 deletions

diff --git a/README.md b/README.md
index 22a7821..c3c77c2 100644
--- a/README.md
+++ b/README.md
@@ -12,7 +12,7 @@ Hexai exposes a simple LLM provider interface and uses OpenAI by default for
 code completion when `OPENAI_API_KEY` is present in the environment.
 
 - Required: set `OPENAI_API_KEY` to your OpenAI API key.
-- Optional: set `OPENAI_MODEL` (default: `gpt-4o-mini`).
+- Optional: set `OPENAI_MODEL`.
 - Optional: set `OPENAI_BASE_URL` to point at a compatible endpoint.
 
 If no key is configured, Hexai will fall back to a basic, local completion.
@@ -20,17 +20,17 @@ If no key is configured, Hexai will fall back to a basic, local completion.
 ## CLI usage and configuration
 
 - Run LSP server over stdio:
-  - `hexai -stdio`
+  - `hexai`
 
 - Completion settings:
-  - `-max-tokens`: maximum tokens for LLM completions (default `500`). If the flag isn’t provided, `HEXAI_MAX_TOKENS` is used when set.
+  - `-max-tokens`: maximum tokens for LLM completions. If the flag isn’t provided, `HEXAI_MAX_TOKENS` is used when set.
   - `-context-mode`: how much additional context to include with completion prompts (If the flag isn’t provided, `HEXAI_CONTEXT_MODE` is used when set). One of:
     - `minimal`: no extra context
     - `window`: include a sliding window around the cursor
-    - `file-on-new-func` (default): include the full file only when defining a new function (cursor before the opening `{`)
+    - `file-on-new-func`ude the full file only when defining a new function (cursor before the opening `{`)
     - `always-full`: always include the full file (may be slower/costly)
-  - `-context-window-lines`: line count for the sliding window when `context-mode=window` (default `120`).
-  - `-max-context-tokens`: budget for additional context tokens (default `2000`). If the flag isn’t provided, `HEXAI_MAX_CONTEXT_TOKENS` is used when set.
+  - `-context-window-lines`: line count for the sliding window when `context-mode=window`.
+  - `-max-context-tokens`: budget for additional context tokens. If the flag isn’t provided, `HEXAI_MAX_CONTEXT_TOKENS` is used when set.
 
 Notes:
 - Token estimation for truncation uses a simple 4 chars/token heuristic.
@@ -38,13 +38,13 @@ Notes:
 
 ### Flags quick reference
 
-| Flag                    | Default            | Env override               | Description                                        |
-|-------------------------|--------------------|----------------------------|----------------------------------------------------|
-| `-stdio`                | `true`             | —                          | Run as LSP over stdio (only supported mode).       |
-| `-log`                  | `/tmp/hexai.log`   | —                          | Path to log file (optional).                       |
-| `-max-tokens`           | `500`              | `HEXAI_MAX_TOKENS`         | Max tokens for LLM completions.                    |
-| `-context-mode`         | `file-on-new-func` | `HEXAI_CONTEXT_MODE`       | `minimal` `window` `file-on-new-func` `always-full` |
-| `-context-window-lines` | `120`              | `HEXAI_CONTEXT_WINDOW_LINES` | Lines around cursor when using `window` mode.      |
-| `-max-context-tokens`   | `2000`             | `HEXAI_MAX_CONTEXT_TOKENS` | Token budget for additional context.               |
-| `-log-preview-limit`    | `0` (unlimited)    | `HEXAI_LOG_PREVIEW_LIMIT`  | Limit characters shown in LLM preview logs.        |
-| `-no-disk-io`           | `true`             | `HEXAI_NO_DISK_IO`         | Disallow any disk reads for context.               |
+| Flag                    | Env override               | Description                                        |
+|-------------------------|----------------------------|----------------------------------------------------|
+| `-stdio`                | —                          | Run as LSP over stdio (only supported mode).       |
+| `-log`                  | —                          | Path to log file (optional).                       |
+| `-max-tokens`           | `HEXAI_MAX_TOKENS`         | Max tokens for LLM completions.                    |
+| `-context-mode`         | `HEXAI_CONTEXT_MODE`       | `minimal` `window` `file-on-new-func` `always-full` |
+| `-context-window-lines` | `HEXAI_CONTEXT_WINDOW_LINES` | Lines around cursor when using `window` mode.      |
+| `-max-context-tokens`   | `HEXAI_MAX_CONTEXT_TOKENS` | Token budget for additional context.               |
+| `-log-preview-limit`    | `HEXAI_LOG_PREVIEW_LIMIT`  | Limit characters shown in LLM preview logs.        |
+| `-no-disk-io`           | `HEXAI_NO_DISK_IO`         | Disallow any disk reads for context.               |