@cldmv/llm-translator 0.3.0 → 0.3.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (2) hide show
  1. package/README.md +4 -4
  2. package/package.json +5 -1
package/README.md CHANGED
@@ -17,14 +17,14 @@ Use it as a library when you want the conversion without any I/O (`request`, `re
17
17
 
18
18
  ## ✨ What's New
19
19
 
20
- ### Latest: v0.3.0 (November 2026)
20
+ ### Latest: v0.3.1 (November 2026)
21
21
 
22
- - **Usage reporting, `route()` and the server `authorize` hook** — every `send()` reply has a `usage` promise (and an `onUsage` callback) with the final token counts; `api.route(body)` says which upstream and model a request would use; `server.start` and `server.plugin` take `authorize(request)` and `onUsage(report, context)` so an app with its own key system can gate and meter requests.
23
- - **More servers and endpoint pools** — drivers for llama.cpp, Ollama, SGLang and the DeepSeek API, failover and load balancing between upstreams that serve the same model, the `reasoningRoundTrip` and `toolSearch` settings, estimated usage when an upstream sends none, and `GET /v1/models/{id}`. Requires Node.js 24 or newer.
24
- - [View full v0.3.0 Changelog](https://github.com/CLDMV/llm-translator/blob/master/docs/changelog/v0/v0.3.0.md)
22
+ - **Patched lint toolchain** — overrides move the dev-only `minimatch` and `katex` (pulled in by ESLint) to versions that fix two advisories with no in-range update, and the v0.3.0 notes now link the right issues. No runtime changes; a drop-in for v0.3.0.
23
+ - [View full v0.3.1 Changelog](https://github.com/CLDMV/llm-translator/blob/master/docs/changelog/v0/v0.3.1.md)
25
24
 
26
25
  ### Recent Releases
27
26
 
27
+ - **v0.3.0** (November 2026) — usage reporting from `send()`, the `route()` leaf and the server `authorize` hook; drivers for llama.cpp, Ollama, SGLang and the DeepSeek API, endpoint pools, `reasoningRoundTrip` and `toolSearch`; requires Node.js 24 ([Changelog](https://github.com/CLDMV/llm-translator/blob/master/docs/changelog/v0/v0.3.0.md))
28
28
  - **v0.2.0** (October 2026) — the first working translator: six Anthropic Messages and OpenAI Chat Completions converters, backend detection (vLLM, OpenAI, Anthropic, generic), model aliases, the optional Fastify server and the `llm-translator` CLI for Claude Code ([Changelog](https://github.com/CLDMV/llm-translator/blob/master/docs/changelog/v0/v0.2.0.md))
29
29
  - **v0.1.0** (October 2026) — package scaffold: ESM and CommonJS entry points, tsup build, the `llm-translator-dev` condition with devcheck, tests on `@cldmv/vitest-runner` and the CLDMV v4 CI/release pipeline; the only export was a placeholder ([Changelog](https://github.com/CLDMV/llm-translator/blob/master/docs/changelog/v0/v0.1.0.md))
30
30
 
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@cldmv/llm-translator",
3
- "version": "0.3.0",
3
+ "version": "0.3.1",
4
4
  "description": "Bidirectional Anthropic Messages ⇄ OpenAI Chat Completions translator with backend auto-detection (vLLM and other OpenAI-compatible servers), built on slothlet, with an optional standalone server for running Claude Code against local models.",
5
5
  "type": "module",
6
6
  "license": "Apache-2.0",
@@ -130,5 +130,9 @@
130
130
  "tsup": "^9.0.0",
131
131
  "typescript": "^6.0.3",
132
132
  "vitest": "^5.0.5"
133
+ },
134
+ "overrides": {
135
+ "katex": ">=0.19.1 <1.0.0",
136
+ "minimatch": "^11.0.3"
133
137
  }
134
138
  }