feat(provider-openai): standalone OpenAI Chat Completions provider worker

.github/workflows/create-tag.yml

@@ -20,6 +20,7 @@ on:
           - iii-lsp-vscode
           - image-resize
           - mcp
+          - provider-openai
           - session-manager
           - shell
           - storage

.github/workflows/release.yml

@@ -14,6 +14,7 @@ on:
       - 'iii-lsp/v*'
       - 'image-resize/v*'
       - 'mcp/v*'
+      - 'provider-openai/v*'
       - 'session-manager/v*'
       - 'shell/v*'
       - 'storage/v*'

README.md

@@ -47,6 +47,7 @@ npx skills add iii-hq/iii --all
 | [`iii-lsp-vscode`](iii-lsp-vscode/) | Node | VS Code extension that embeds `iii-lsp`. |
 | [`image-resize`](image-resize/) | Rust | Image resize via channel I/O — JPEG/PNG/WebP with EXIF auto-orient, scale-to-fit / crop-to-fit. |
 | [`mcp`](mcp/) | Rust | MCP 2025-06-18 Streamable HTTP bridge — exposes iii functions tagged `mcp.expose` as MCP tools. |
+| [`provider-openai`](provider-openai/) | Rust | OpenAI Chat Completions provider behind `llm-router` — `provider::openai::stream` with reasoning support and live chat-model discovery. |
 | [`shell`](shell/) | Rust | Unix shell + filesystem worker — `shell::exec` with allowlist/denylist/timeout/output caps and background jobs; `fs::ls`/`stat`/`mkdir`/`rm`/`chmod`/`mv`/`grep`/`sed`/`read`/`write` with host jail, denylist, and size caps. |
 | [`storage`](storage/) | Rust | S3-compatible object storage across AWS S3, GCS, Cloudflare R2, and a managed local rustfs backend. Streamed uploads, presigned URLs, and object change triggers. |
 | [`todo-worker`](todo-worker/) | Node | Quickstart CRUD todo worker using the Node iii SDK. |

llm-router/README.md

@@ -171,6 +171,8 @@ A provider worker must:
 The first real provider implementing this protocol is
 [`provider-anthropic/`](../provider-anthropic/) — useful as a reference
 implementation alongside the scripted provider in the integration tests.
+[`provider-openai/`](../provider-openai/) follows the same structure for the
+OpenAI Chat Completions API (native structured output, reasoning_effort).
 
 ## Local development & testing
 

provider-openai/.gitignore

@@ -0,0 +1 @@
+/target