feat: audio surfaces (TTS + transcription), imagegen.Editor, llamaswap health probe

- New leaf package `audio` (ADR-0017): SpeechModel/SpeechProvider and
  TranscriptionModel/TranscriptionProvider with imagegen conventions
  (zero value = backend default, functional options + Apply, bytes
  in/out, never URLs). Root re-exports added.
- imagegen.Editor (ADR-0018): optional image-to-image interface —
  EditRequest carries the generation knobs plus Init image and
  denoising Strength; separate interface so existing Models keep
  compiling.
- provider/llamaswap implements all of it: POST /v1/audio/speech (JSON,
  raw-audio response, MIME from Content-Type with format fallback),
  POST /v1/audio/transcriptions (multipart, response_format=json),
  ListVoices (GET /v1/audio/voices?model=, tolerant of string-list and
  object-list shapes), POST /sdapi/v1/img2img (txt2img wire +
  init_images/denoising_strength, shared image decode), and Health(ctx)
  (GET /health) — a cheap liveness probe for often-offline hosts.
- Hermetic httptest coverage for every new wire shape and validation
  path; README sections + support-matrix footnote updated in the same
  commit (also corrects the stale /v1/images/generations claim — the
  image path has been SDAPI since the seed fix).

First consumer: mort's llamaswap media tool cluster (status / image /
TTS / STT agent tools against the netherstorm host).

Co-Authored-By: Claude Fable 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01AXQxVhXBw8PwFAtsVrXSmj

provider/llamaswap/image.go

@@ -85,8 +85,13 @@ func (m *imageModel) Generate(ctx context.Context, req imagegen.Request, opts ..
 	if err := m.p.doJSON(ctx, http.MethodPost, "/sdapi/v1/txt2img", m.id, &wire, &resp); err != nil {
 		return nil, err
 	}
+	return decodeImages(m.p.name, m.id, &resp)
+}
 
-	out := &imagegen.Result{Raw: &resp}
+// decodeImages converts an SDAPI response's base64 images into an
+// imagegen.Result, erroring when nothing decodable came back.
+func decodeImages(provider, model string, resp *txt2imgResponse) (*imagegen.Result, error) {
+	out := &imagegen.Result{Raw: resp}
 	for i, b64 := range resp.Images {
 		if b64 == "" {
 			continue
@@ -99,14 +104,70 @@ func (m *imageModel) Generate(ctx context.Context, req imagegen.Request, opts ..
 	}
 	if len(out.Images) == 0 {
 		return nil, &llm.APIError{
-			Provider: m.p.name,
-			Model:    m.id,
+			Provider: provider,
+			Model:    model,
 			Message:  "image response contained no images",
 		}
 	}
 	return out, nil
 }
 
+// img2imgRequest is the stable-diffusion.cpp sd-server A1111 request shape
+// (POST /sdapi/v1/img2img): txt2img's fields plus the init image(s) and
+// denoising strength. Same endpoint-family choice as txt2img — the OpenAI
+// /v1/images/edits route is multipart and drops `seed` on this sd-server
+// build, while the SDAPI shape reuses doJSON and keeps seed parity.
+type img2imgRequest struct {
+	txt2imgRequest
+	InitImages        []string `json:"init_images"`
+	DenoisingStrength *float64 `json:"denoising_strength,omitempty"`
+}
+
+// Edit implements imagegen.Editor via POST {base}/sdapi/v1/img2img.
+func (m *imageModel) Edit(ctx context.Context, req imagegen.EditRequest, opts ...imagegen.EditOption) (*imagegen.Result, error) {
+	req = req.Apply(opts...)
+	if strings.TrimSpace(req.Prompt) == "" {
+		return nil, fmt.Errorf("%w: image edit requires a prompt", llm.ErrUnsupported)
+	}
+	if len(req.Init.Data) == 0 {
+		return nil, fmt.Errorf("%w: image edit requires an init image", llm.ErrUnsupported)
+	}
+	if req.N < 0 {
+		return nil, fmt.Errorf("%w: image count N must be >= 0, got %d", llm.ErrUnsupported, req.N)
+	}
+	if req.Strength != nil && (*req.Strength < 0 || *req.Strength > 1) {
+		return nil, fmt.Errorf("%w: edit strength must be in [0,1], got %g", llm.ErrUnsupported, *req.Strength)
+	}
+
+	width, height, err := parseSize(req.Size)
+	if err != nil {
+		return nil, fmt.Errorf("%w: %v", llm.ErrUnsupported, err)
+	}
+
+	wire := img2imgRequest{
+		txt2imgRequest: txt2imgRequest{
+			Model:          m.id,
+			Prompt:         req.Prompt,
+			NegativePrompt: req.NegativePrompt,
+			Seed:           req.Seed,
+			Steps:          req.Steps,
+			CFGScale:       req.CFGScale,
+			Width:          width,
+			Height:         height,
+			SampleMethod:   req.Sampler,
+			BatchCount:     req.N,
+		},
+		InitImages:        []string{base64.StdEncoding.EncodeToString(req.Init.Data)},
+		DenoisingStrength: req.Strength,
+	}
+
+	var resp txt2imgResponse
+	if err := m.p.doJSON(ctx, http.MethodPost, "/sdapi/v1/img2img", m.id, &wire, &resp); err != nil {
+		return nil, err
+	}
+	return decodeImages(m.p.name, m.id, &resp)
+}
+
 // parseSize splits a "WxH" string into width/height pointers. "" yields
 // (nil, nil) so the model's own default resolution applies.
 func parseSize(size string) (*int, *int, error) {