add support set up for Windsurf and Antigravity

AGENTS.md

@@ -45,7 +45,7 @@ pantry store \
 
 Categories: `decision`, `bug`, `pattern`, `context`, `learning`.
 
-Set `--source` to your agent identifier: `claude-code`, `codex`, `cursor`, or `opencode`.
+Set `--source` to your agent identifier: `claude-code`, `codex`, `cursor`, `windsurf`, `antigravity`, or `opencode`.
 
 `--project` defaults to the current directory name — only set it explicitly if storing a note for a different project.
 
@@ -72,7 +72,7 @@ You MUST store a note when any of these happen:
 Run once to auto-install hooks for your agent:
 
 ```bash
-pantry setup cursor   # or: claude-code, codex, opencode
+pantry setup cursor   # or: claude-code, windsurf, antigravity, codex, opencode
 ```
 
 To remove: `pantry uninstall cursor`

README.md

@@ -4,7 +4,7 @@ Local note storage for coding agents. Your agent keeps notes on decisions, bugs,
 
 ## Features
 
-- **Works with multiple agents** — Claude Code, Cursor, Codex, OpenCode, RooCode. One command sets up MCP config for your agent.
+- **Works with multiple agents** — Claude Code, Cursor, Windsurf, Antigravity, Codex, OpenCode, RooCode. One command sets up MCP config for your agent.
 - **MCP native** — Runs as an MCP server exposing `pantry_store`, `pantry_search`, and `pantry_context` as tools.
 - **Local-first** — Everything stays on your machine. Notes are stored as Markdown in `~/.pantry/shelves/`, readable in Obsidian or any editor.
 - **Zero idle cost** — No background processes, no daemon, no RAM overhead. The MCP server only runs when the agent starts it.
@@ -44,7 +44,7 @@ pantry init
 ### Connect your agent
 
 ```bash
-pantry setup claude-code   # or: cursor, codex, opencode, roocode
+pantry setup claude-code   # or: cursor, windsurf, antigravity, codex, opencode, roocode
 ```
 
 This writes the MCP server entry into your agent's config file. Restart the agent and pantry will be available as a tool.
@@ -53,7 +53,7 @@ Run `pantry doctor` to verify everything is working.
 
 ### Tell your agent to use Pantry
 
-MCP registration makes the tools available, but your agent also needs instructions to actually use them. The `setup` command installs a skill file automatically for agents that support it (Claude Code, Cursor, Codex). For other agents — or if you prefer to use a project-level rules file — add the following to your `AGENTS.md`, `.rules`, `CLAUDE.md`, or equivalent:
+MCP registration makes the tools available, but your agent also needs instructions to actually use them. The `setup` command installs a skill file automatically for agents that support it (Claude Code, Cursor, Windsurf, Antigravity, Codex). For other agents — or if you prefer to use a project-level rules file — add the following to your `AGENTS.md`, `.rules`, `CLAUDE.md`, or equivalent:
 
 ```markdown
 ## Pantry — persistent notes
@@ -80,15 +80,18 @@ Do not skip either step. Notes are how context survives across sessions.
 Keyword search (FTS5) works with no extra setup. To also enable semantic vector search, configure an embedding provider in `~/.pantry/config.yaml`:
 
 **Ollama (local, free):**
+
 ```yaml
 embedding:
   provider: ollama
   model: nomic-embed-text
   base_url: http://localhost:11434
 ```
+
 Install [Ollama](https://ollama.com), then: `ollama pull nomic-embed-text`
 
 **OpenAI:**
+
 ```yaml
 embedding:
   provider: openai
@@ -97,14 +100,25 @@ embedding:
 ```
 
 **OpenRouter:**
+
 ```yaml
 embedding:
   provider: openrouter
   model: openai/text-embedding-3-small
   api_key: sk-or-...
 ```
 
+**Google (Gemini API):**
+
+```yaml
+embedding:
+  provider: google
+  model: gemini-embedding-001
+  api_key: AIzaSy...
+```
+
 After changing providers, rebuild the vector index:
+
 ```bash
 pantry reindex
 ```
@@ -116,9 +130,9 @@ All config file values can be overridden with environment variables. They take p
 | Variable | Description | Example |
 |----------|-------------|---------|
 | `PANTRY_HOME` | Override pantry home directory | `/data/pantry` |
-| `PANTRY_EMBEDDING_PROVIDER` | Embedding provider | `ollama`, `openai`, `openrouter` |
-| `PANTRY_EMBEDDING_MODEL` | Embedding model name | `text-embedding-3-small` |
-| `PANTRY_EMBEDDING_API_KEY` | API key for the embedding provider | `sk-...` |
+| `PANTRY_EMBEDDING_PROVIDER` | Embedding provider | `ollama`, `openai`, `openrouter`, `google` |
+| `PANTRY_EMBEDDING_MODEL` | Embedding model name | `text-embedding-3-small`, `gemini-embedding-001` |
+| `PANTRY_EMBEDDING_API_KEY` | API key for the embedding provider | `sk-...`, `AIzaSy...` |
 | `PANTRY_EMBEDDING_BASE_URL` | Base URL for the embedding API | `http://localhost:11434` |
 | `PANTRY_CONTEXT_SEMANTIC` | Semantic search mode | `auto`, `always`, `never` |
 

internal/config/config.go

@@ -113,9 +113,9 @@ func LoadConfig(path string) (*Config, error) {
 // Validate returns an error if the configuration contains invalid values.
 // Call this after LoadConfig to surface misconfiguration at startup.
 func (c *Config) Validate() error {
-	validProviders := map[string]bool{"ollama": true, "openai": true, "openrouter": true}
+	validProviders := map[string]bool{"ollama": true, "openai": true, "openrouter": true, "google": true}
 	if !validProviders[c.Embedding.Provider] {
-		return fmt.Errorf("invalid embedding.provider %q: must be one of ollama, openai, openrouter", c.Embedding.Provider)
+		return fmt.Errorf("invalid embedding.provider %q: must be one of ollama, openai, openrouter, google", c.Embedding.Provider)
 	}
 
 	if c.Embedding.Model == "" {
@@ -127,7 +127,7 @@ func (c *Config) Validate() error {
 		return fmt.Errorf("invalid context.semantic %q: must be one of auto, always, never", c.Context.Semantic)
 	}
 
-	if c.Embedding.Provider == "openai" || c.Embedding.Provider == "openrouter" {
+	if c.Embedding.Provider == "openai" || c.Embedding.Provider == "openrouter" || c.Embedding.Provider == "google" {
 		if c.Embedding.APIKey == nil || *c.Embedding.APIKey == "" {
 			return fmt.Errorf("embedding.api_key is required for provider %q", c.Embedding.Provider)
 		}
@@ -162,10 +162,10 @@ func GetDefaultConfigTemplate() string {
 # Embedding provider for semantic search.
 # Without this, keyword search (FTS5) still works.
 embedding:
-  provider: ollama              # ollama | openai | openrouter
+  provider: ollama              # ollama | openai | openrouter | google
   model: nomic-embed-text
   base_url: http://localhost:11434
-  # api_key: sk-...            # required for openai/openrouter
+  # api_key: sk-...            # required for openai/openrouter/google
 
 # How items are retrieved at session start.
 # "auto" uses vectors when available, falls back to keywords.

internal/embeddings/factory.go

@@ -43,6 +43,18 @@ func NewProvider(cfg config.EmbeddingConfig) (Provider, error) {
 
 		return NewOpenAIProvider(cfg.Model, *cfg.APIKey, baseURL), nil
 
+	case "google":
+		if cfg.APIKey == nil || *cfg.APIKey == "" {
+			return nil, errors.New("API key required for Google provider")
+		}
+
+		baseURL := ""
+		if cfg.BaseURL != nil {
+			baseURL = *cfg.BaseURL
+		}
+
+		return NewGoogleProvider(cfg.Model, *cfg.APIKey, baseURL), nil
+
 	default:
 		return nil, fmt.Errorf("unknown embedding provider: %s", cfg.Provider)
 	}

internal/embeddings/google.go

@@ -0,0 +1,106 @@
+package embeddings
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"strings"
+)
+
+// GoogleProvider implements embedding generation using Google's Gemini API.
+type GoogleProvider struct {
+	model   string
+	apiKey  string
+	baseURL string
+	client  *http.Client
+}
+
+// NewGoogleProvider creates a new Google Gemini embedding provider.
+func NewGoogleProvider(model string, apiKey string, baseURL string) *GoogleProvider {
+	if baseURL == "" {
+		baseURL = "https://generativelanguage.googleapis.com/v1beta"
+	}
+
+	return &GoogleProvider{
+		model:   model,
+		apiKey:  apiKey,
+		baseURL: strings.TrimSuffix(baseURL, "/"),
+		client:  &http.Client{},
+	}
+}
+
+type googleEmbedRequest struct {
+	Model   string             `json:"model"`
+	Content googleEmbedContent `json:"content"`
+}
+
+type googleEmbedContent struct {
+	Parts []googleEmbedPart `json:"parts"`
+}
+
+type googleEmbedPart struct {
+	Text string `json:"text"`
+}
+
+type googleEmbedResponse struct {
+	Embedding struct {
+		Values []float32 `json:"values"`
+	} `json:"embedding"`
+}
+
+// Embed generates an embedding vector using Google Gemini API.
+func (p *GoogleProvider) Embed(ctx context.Context, text string) ([]float32, error) {
+	// The Gemini API requires the model name to be prefixed with "models/"
+	modelPath := p.model
+	if !strings.HasPrefix(modelPath, "models/") {
+		modelPath = "models/" + modelPath
+	}
+
+	url := fmt.Sprintf("%s/%s:embedContent", p.baseURL, modelPath)
+
+	reqData := googleEmbedRequest{
+		Model: modelPath,
+		Content: googleEmbedContent{
+			Parts: []googleEmbedPart{{Text: text}},
+		},
+	}
+
+	jsonData, err := json.Marshal(reqData)
+	if err != nil {
+		return nil, fmt.Errorf("failed to marshal request: %w", err)
+	}
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, url, bytes.NewReader(jsonData))
+	if err != nil {
+		return nil, fmt.Errorf("failed to create request: %w", err)
+	}
+
+	req.Header.Set("Content-Type", "application/json")
+	req.Header.Set("x-goog-api-key", p.apiKey)
+
+	resp, err := p.client.Do(req)
+	if err != nil {
+		return nil, fmt.Errorf("failed to call Google API: %w", err)
+	}
+
+	defer func() { _ = resp.Body.Close() }()
+
+	if resp.StatusCode != http.StatusOK {
+		body, _ := io.ReadAll(resp.Body)
+		return nil, fmt.Errorf("google API returned status %d: %s", resp.StatusCode, string(body))
+	}
+
+	var response googleEmbedResponse
+	if err := json.NewDecoder(resp.Body).Decode(&response); err != nil {
+		return nil, fmt.Errorf("failed to decode response: %w", err)
+	}
+
+	if len(response.Embedding.Values) == 0 {
+		return nil, fmt.Errorf("no embedding values returned by google api")
+	}
+
+	return response.Embedding.Values, nil
+}