Add GPT 6 Astra and Claude Fable 5.1; refresh Go dependencies

.github/workflows/ci.yml

@@ -22,9 +22,9 @@ jobs:
         run: go build ./...
 
       - name: Lint
-        uses: golangci/golangci-lint-action@v6
-        with:
-          version: latest
+        # Build with the project's Go toolchain; the v1 release binary was
+        # built with Go 1.24 and cannot analyze this module's Go 1.25 target.
+        run: go run github.com/golangci/golangci-lint/cmd/golangci-lint@v1.64.8 run
 
       - name: Vet
         run: go vet ./...

AGENTS.md

@@ -9,7 +9,7 @@ standalone on a local repository or connected to QualityMax. Through `/orch`
 it can launch Claude Code, Codex, or OpenCode while keeping qmax tools and
 terminal UX — or run the built-in loop on Anthropic, Cerebras, or Ollama.
 
-This repository is **Go 1.24+**, not TypeScript. The thing users install is one
+This repository is **Go 1.25.13+**, not TypeScript. The thing users install is one
 compiled binary.
 
 ## Positioning

CLAUDE.md

@@ -1,6 +1,6 @@
 # CLAUDE.md
 
-qmax-code is a **Go** terminal **QA** agent (Go 1.24+). It ships as one
+qmax-code is a **Go** terminal **QA** agent (Go 1.25.13+). It ships as one
 compiled binary: `curl | bash` installs a file, not a Node or Python runtime.
 Claude Code, Codex, and OpenCode are separate subprocesses that `/orch` can
 host; they are not this repository's runtime.

CONTRIBUTING.md

@@ -26,7 +26,7 @@ Thanks for your interest in improving `qmax-code`. This document covers everythi
 
 | Requirement | Notes |
 |---|---|
-| **Go 1.24+** | See `go.mod` for the exact version. `go version` to check. |
+| **Go 1.25.13+** | See `go.mod` for the exact version. `go version` to check. |
 | **Inference backend** | The direct API path needs an Anthropic key. You can instead develop against a logged-in Claude Code/Codex CLI, an enabled OpenCode provider, Cerebras, or Ollama. |
 | **QualityMax account** | Required only for connected cloud tools (test generation, crawl, repo review). Standalone `--local` development and most unit tests do not need one. |
 

README.md

@@ -197,7 +197,7 @@ go build -o qmax-code .
 ./qmax-code --version
 ```
 
-Go 1.24 or newer is required for source builds.
+Go 1.25.13 or newer is required for source builds.
 
 ## Why Go
 
@@ -334,7 +334,7 @@ troubleshooting.
 | Backend | Select with | Authentication | Notes |
 | --- | --- | --- | --- |
 | Anthropic API | `/api` or `/orch` | `ANTHROPIC_API_KEY` or OS keychain | Built-in agent loop; tool set follows connected vs. standalone mode. |
-| Claude Code | `/cc` or `/orch` | Local Claude Code login | CLI subprocess; qmax tools arrive through MCP. Agent SDK usage may be separately metered by Anthropic. |
+| Claude Code | `/cc` or `/orch` | Local Claude Code login | CLI subprocess; qmax tools arrive through MCP. Uses Claude Code authentication and plan/provider billing; see [subscription setup](docs/ORCHESTRATION.md#claude-code-subscription-billing). |
 | Codex | `/codex` or `/orch` | Local Codex login | CLI subprocess using the user's OpenAI access; qmax tools arrive through MCP. |
 | Cerebras | `/gemma`, `/orch`, or `--backend cerebras` | `CEREBRAS_API_KEY` or OS keychain | Built-in native function calling. Fast inference (~1000–2000+ tok/s): GPT-OSS 120B, GLM 4.7, Gemma 4 (vision + effort). **Qwen 3.8 coming soon.** |
 | OpenCode | `/opencode` or `/orch` | Per-provider key in OS keychain | CLI subprocess for opt-in Z.AI, Groq, and OpenRouter providers. |

codexrunner/model.go

@@ -14,6 +14,7 @@ var (
 	ErrInvalidModel = errors.New("codex runner: invalid model")
 
 	supportedModels = [...]string{
+		"gpt-6-astra",
 		"gpt-5.6-sol",
 		DefaultModel,
 		"gpt-5.6-luna",

docs/ORCHESTRATION.md

@@ -56,6 +56,86 @@ QualityMax authentication is separate from backend authentication. Use
 `--local` to skip QualityMax authentication entirely. QualityMax cloud tools
 require connected mode and `qmax-code login`.
 
+## Go integration with Claude Code and Codex
+
+qmax-code calls both native harnesses directly from Go. As of September 5,
+2026, their official harness SDKs are available for Python and TypeScript;
+neither provider documents an official Go harness SDK.
+
+| Harness | Go integration used here | Continuation |
+| --- | --- | --- |
+| Claude Code | `os/exec` with `claude --print --output-format stream-json` | `--resume` with the native session ID |
+| Codex | The Go `codexrunner` package with `codex exec --json` | `codex exec resume` with the native thread ID |
+
+Anthropic explicitly recommends a CLI subprocess for other languages in its
+[Agent SDK overview](https://code.claude.com/docs/en/agent-sdk). OpenAI documents
+the JSONL stream and saved CLI authentication in its
+[non-interactive guide](https://learn.chatgpt.com/docs/non-interactive-mode);
+its [SDK guide](https://developers.openai.com/codex/sdk/) covers the Python and
+TypeScript libraries. Codex's [app server](https://learn.chatgpt.com/docs/app-server)
+also exposes a language-independent JSON-RPC interface for richer custom clients.
+
+The providers' Go **API client** SDKs are separate from these coding harnesses.
+Adding an API client would not provide the native agent loop or subscription
+login. The current subprocess integration keeps qmax-code a single Go binary,
+with Claude Code and Codex installed and updated separately. `go mod tidy` or
+updating Go modules does not update those CLI executables.
+
+## GPT 6 Astra and Claude Fable 5.1
+
+Select **GPT 6 Astra** under Codex, or **Fable 5.1** under Claude Code or
+Anthropic API in `/orch`. The exact IDs are `gpt-6-astra` and
+`claude-fable-5-1`; `fable` is a shorthand for Fable 5.1. Existing defaults
+and the explicit Fable 5 model remain available.
+
+```bash
+qmax-code --local --backend codex --model gpt-6-astra
+qmax-code --local --backend cc --model fable
+# Direct API billing:
+qmax-code --local --backend api --model claude-fable-5-1
+```
+
+The `/orch` Codex selection is saved separately from Claude's preference and
+survives `/clear`. **Codex default** (or `--backend codex --model auto`)
+uses Codex's own model configuration. Approval and sandbox policy continue to
+come from Codex configuration. Model access depends on the account used by
+the native CLI. See the [Astra model reference](https://developers.openai.com/api/docs/models/gpt-6-astra)
+and [Fable 5.1 reference](https://platform.claude.com/docs/en/models/fable-5-1/overview).
+
+## Claude Code subscription billing
+
+Yes: qmax-code can use the native Claude Code harness with a Claude subscription
+login, without an Anthropic API key. The `cc` backend runs `claude --print`
+and connects qmax tools through MCP. It inherits Claude Code authentication;
+it does not require `qmax-code cc connect` or a QualityMax login in local mode.
+
+For subscription authentication, remove the API-key override from the shell
+that launches qmax-code, then open Claude Code and use `/login` to sign in with
+your Claude account and `/status` to verify authentication:
+
+```bash
+unset ANTHROPIC_API_KEY
+claude
+# After signing in and exiting Claude Code:
+qmax-code --local --backend cc --model fable
+```
+
+An inherited `ANTHROPIC_API_KEY` takes precedence over subscription login and
+causes API billing. See [Anthropic's authentication guidance](https://support.claude.com/en/articles/12304248-manage-api-key-environment-variables-in-claude-code).
+
+As of September 5, 2026, the separate third-party Agent SDK billing change
+announced for June 15 remains paused. Anthropic's
+[June 15 update](https://support.claude.com/en/articles/15036540-use-the-claude-agent-sdk-with-your-claude-plan)
+confirms that Agent SDK, `claude -p`, and third-party app usage still draw from
+subscription limits. Conductor also documents continued subscription use in
+its [billing update](https://www.conductor.build/blog/claude-subscription-update).
+
+Fable 5.1 specifically requires Claude Code **2.1.255 or later**. Max and eligible
+premium Team/legacy Enterprise seats include Fable usage up to 50% of the weekly
+limit. Pro and standard seats use paid usage credits for Fable 5.1 from the first
+request. A subscription login therefore does not guarantee included Fable usage;
+check your tier and extra-usage settings in [Anthropic's Fable plan guide](https://support.claude.com/en/articles/15424964-claude-fable-models-on-your-plan).
+
 ## Standalone local-only orchestration
 
 Every backend can be selected in standalone mode:

go.mod

@@ -1,44 +1,47 @@
 module github.com/qualitymax/qmax-code
 
-go 1.24.0
+go 1.25.13
 
 require (
 	github.com/BourgeoisBear/rasterm v1.1.2
 	github.com/Quality-Max/qmax-receipt v0.1.0
 	github.com/charmbracelet/bubbletea v1.3.10
-	github.com/charmbracelet/glamour v0.8.0
-	github.com/charmbracelet/lipgloss v1.1.0
-	github.com/charmbracelet/x/ansi v0.10.1
+	github.com/charmbracelet/glamour v1.0.0
+	github.com/charmbracelet/lipgloss v1.1.1-0.20250404203927-76690c660834
+	github.com/charmbracelet/x/ansi v0.11.8
 	github.com/chzyer/readline v1.5.1
-	github.com/coder/websocket v1.8.14
-	github.com/getsentry/sentry-go v0.44.1
-	golang.org/x/sys v0.36.0
-	golang.org/x/term v0.22.0
+	github.com/coder/websocket v1.8.15
+	github.com/getsentry/sentry-go v0.49.0
+	golang.org/x/sys v0.47.0
+	golang.org/x/term v0.45.0
 )
 
 require (
-	github.com/alecthomas/chroma/v2 v2.14.0 // indirect
+	github.com/alecthomas/chroma/v2 v2.27.0 // indirect
 	github.com/aymanbagabas/go-osc52/v2 v2.0.1 // indirect
 	github.com/aymerick/douceur v0.2.0 // indirect
-	github.com/charmbracelet/colorprofile v0.2.3-0.20250311203215-f60798e515dc // indirect
-	github.com/charmbracelet/x/cellbuf v0.0.13-0.20250311204145-2c3ea96c31dd // indirect
-	github.com/charmbracelet/x/term v0.2.1 // indirect
-	github.com/dlclark/regexp2 v1.11.0 // indirect
+	github.com/charmbracelet/colorprofile v0.4.3 // indirect
+	github.com/charmbracelet/x/cellbuf v0.0.15 // indirect
+	github.com/charmbracelet/x/exp/slice v0.0.0-20260902165432-6f6ad8b37b0a // indirect
+	github.com/charmbracelet/x/term v0.2.2 // indirect
+	github.com/clipperhouse/displaywidth v0.11.0 // indirect
+	github.com/clipperhouse/uax29/v2 v2.7.0 // indirect
+	github.com/dlclark/regexp2/v2 v2.7.1 // indirect
 	github.com/erikgeiser/coninput v0.0.0-20211004153227-1c3628e74d0f // indirect
 	github.com/gorilla/css v1.0.1 // indirect
-	github.com/lucasb-eyer/go-colorful v1.2.0 // indirect
-	github.com/mattn/go-isatty v0.0.20 // indirect
+	github.com/lucasb-eyer/go-colorful v1.4.1 // indirect
+	github.com/mattn/go-isatty v0.0.24 // indirect
 	github.com/mattn/go-localereader v0.0.1 // indirect
-	github.com/mattn/go-runewidth v0.0.16 // indirect
+	github.com/mattn/go-runewidth v0.0.29 // indirect
 	github.com/microcosm-cc/bluemonday v1.0.27 // indirect
 	github.com/muesli/ansi v0.0.0-20230316100256-276c6243b2f6 // indirect
 	github.com/muesli/cancelreader v0.2.2 // indirect
 	github.com/muesli/reflow v0.3.0 // indirect
 	github.com/muesli/termenv v0.16.0 // indirect
 	github.com/rivo/uniseg v0.4.7 // indirect
-	github.com/xo/terminfo v0.0.0-20220910002029-abceb7e1c41e // indirect
-	github.com/yuin/goldmark v1.7.4 // indirect
-	github.com/yuin/goldmark-emoji v1.0.3 // indirect
-	golang.org/x/net v0.27.0 // indirect
-	golang.org/x/text v0.16.0 // indirect
+	github.com/xo/terminfo v1.0.0 // indirect
+	github.com/yuin/goldmark v1.8.6 // indirect
+	github.com/yuin/goldmark-emoji v1.0.6 // indirect
+	golang.org/x/net v0.58.0 // indirect
+	golang.org/x/text v0.41.0 // indirect
 )