@loongphy/codex-auth-linux-x64 0.2.0 → 0.2.2-alpha.1

package/README.md

@@ -67,6 +67,7 @@ codex-auth list # list all accounts
 codex-auth login # run `codex login`, then add the current account
 codex-auth switch [<email>] # switch active account (interactive or partial/fragment match)
 codex-auth import <path> [--alias <alias>] # smart import: file -> single import, folder -> batch import
+codex-auth import --cpa [<path>] # import CPA flat token JSON from one file or directory
 codex-auth import --purge [<path>] # rebuild registry.json from auth files for the current version
 codex-auth remove # remove accounts (interactive multi-select)
 codex-auth status # show auto-switch/service/api usage status
@@ -103,6 +104,26 @@ Batch import from a folder:
 codex-auth import /path/to/auth-exports
 ```
 
+Import CPA exports directly:
+
+```shell
+codex-auth import --cpa                  # scan ~/.cli-proxy-api/*.json
+codex-auth import --cpa /path/to/cpa-dir
+codex-auth import --cpa /path/to/cpa.json --alias personal
+```
+
+Typical batch import output:
+
+```text
+Scanning /path/to/auth-exports...
+  ✓ imported  token_ryan.taylor.alpha@email.com
+  ✓ updated   token_jane.smith.alpha@email.com
+  ✗ skipped   token_invalid: MalformedJson
+Import Summary: 1 imported, 1 updated, 1 skipped (total 3 files)
+```
+
+`stdout` carries scanning/success/summary lines. Skipped files and warnings stay on `stderr`.
+
 Rebuild `registry.json` from imported auth files:
 
 ```shell
@@ -177,15 +198,82 @@ Changing thresholds updates `registry.json`; Linux/WSL and Windows pick them up
 Changing `config api` updates `registry.json` immediately; `api enable` means API-only and `api disable` means local-sessions-only.
 `codex-auth help` also shows whether auto-switching and usage API calls are currently enabled.
 
+## Q&A
+
+### Why is my usage limit not refreshing?
+
+If `codex-auth` is using local-only usage refresh, it reads the newest `~/.codex/sessions/**/rollout-*.jsonl` file. Recent Codex builds often write `token_count` events with `rate_limits: null`. The local files may still contain older usable usage limit data, but in practice they can lag by several hours, so local-only refresh may show a usage limit snapshot from hours ago instead of your latest state.
+
+- Upstream Codex issue: [openai/codex#14880](https://github.com/openai/codex/issues/14880)
+
+You can switch usage limit refresh to the usage API with:
+
+```shell
+codex-auth config api enable
+```
+
+Then confirm the current mode with:
+
+```shell
+codex-auth status
+```
+
+`status` should show `usage: api`.
+
+Upgrade notes:
+
+- If you are upgrading from `v0.1.x` to the latest `v0.2.x`, API usage refresh is enabled by default.
+- If you previously used an early `v0.2` prerelease/test build and `status` still shows `usage: local`, run `codex-auth config api enable` once to switch to API mode.
+
+### How to import tokens from cli-proxy-api?
+
+If you have token files from `~/.cli-proxy-api/token*.json`, this repository includes a helper script that can convert them into a format codex-auth can read.
+
+The CLI can also import the flat cli-proxy-api / CPA JSON files directly:
+
+```shell
+codex-auth import --cpa                  # default source: ~/.cli-proxy-api
+codex-auth import --cpa /path/to/cpa-dir # scans direct child .json files
+```
+
+Each CPA file is converted in memory to the standard auth snapshot shape before it is written into `~/.codex/accounts/`. Missing or empty `refresh_token` values are skipped as `MissingRefreshToken`.
+
+The script is not bundled in the published npm package, so run it from a clone of this repository:
+
+```shell
+# Convert: ~/.cli-proxy-api → /tmp/tokens
+python3 scripts/convert_tokens.sh
+
+# Or specify custom directories
+python3 scripts/convert_tokens.sh <source_dir> <output_dir>
+```
+
+Then import and switch:
+
+```shell
+codex-auth import /tmp/tokens/
+# or import the CPA files directly without a conversion step
+codex-auth import --cpa
+codex-auth switch
+```
+
+Verify with:
+
+```shell
+codex exec "say hello"
+```
+
+> **Output format:** Each converted file contains `auth_mode`, `OPENAI_API_KEY`, `tokens` (with `id_token`, `access_token`, `refresh_token`, `account_id`), and `last_refresh`. The `refresh_token` field is required for token refresh to work.
+
 ## Disclaimer
 
 This project is provided as-is and use is at your own risk.
 
 **Usage Data Refresh Source:**
-`codex-auth` supports two sources for refreshing account usage/quota information:
+`codex-auth` supports two sources for refreshing account usage/usage limit information:
 
-1. **Local (default):** Scans local `~/.codex/sessions/*/rollout-*.jsonl` files. This mode is strictly local.
-2. **API:** When `config api enable` is on, the tool will make direct HTTPS requests to OpenAI's endpoints using your account's access token.
+1. **API (default):** When `config api enable` is on, the tool makes direct HTTPS requests to OpenAI's endpoints using your account's access token. This is the current default mode.
+2. **Local-only:** When `config api disable` is on, the tool scans local `~/.codex/sessions/*/rollout-*.jsonl` files without making API calls. This mode is safer, but it can be less accurate because recent Codex rollout files often contain `rate_limits: null`, so the latest local usage limit data may lag by several hours.
 
 **API Call Declaration:**
 By enabling API-based usage refresh, this tool will send your ChatGPT access token to OpenAI's servers (specifically `https://chatgpt.com/backend-api/wham/usage`) to fetch current quota information. This behavior may be detected by OpenAI and could violate their terms of service, potentially leading to account suspension or other risks. The decision to use this feature and any resulting consequences are entirely yours.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@loongphy/codex-auth-linux-x64",
-  "version": "0.2.0",
+  "version": "0.2.2-alpha.1",
   "description": "@loongphy/codex-auth binary for linux x64",
   "license": "MIT",
   "repository": {