@copilotkit/pathfinder 1.1.0 → 1.4.0

package/CHANGELOG.md

@@ -1,4 +1,22 @@
-# Changelog
+# @copilotkit/pathfinder
+
+## 1.4.0
+
+### Minor Changes
+
+- Add HTML source type for indexing static HTML documentation sites (cheerio-based parser)
+- Content container auto-detection (main, article, [role="main"], .content, #content)
+- Heading-boundary chunking with headingPath tracking (h1-h3)
+- Code block preservation, list formatting, table formatting in HTML extraction
+- Add pathfinder-docs.yaml for dogfooding Pathfinder on its own documentation
+
+### Patch Changes
+
+- Generalize smoke test script for any Pathfinder instance
+- Add mobile hamburger nav menu to all docs pages
+- Simplify README to match aimock style, add npm metadata (repository, homepage, keywords)
+- Fix Dockerfile to copy pathfinder.yaml for production deploy
+- Fix schema migration: remove version index from generateSchema (was failing on existing databases)
 
 ## 1.1.0
 

package/README.md

@@ -1,284 +1,96 @@
-# Pathfinder
+# pathfinder [![npm version](https://img.shields.io/npm/v/@copilotkit/pathfinder)](https://www.npmjs.com/package/@copilotkit/pathfinder)
 
-A self-hosted MCP server that provides semantic search over your documentation and code. Configure it with a YAML file, deploy with Docker, and give your AI coding agents instant access to your project's knowledge.
-
-## How It Works
-
-Pathfinder indexes your GitHub repositories — documentation (Markdown/MDX) and source code — into a PostgreSQL vector database using OpenAI embeddings. It exposes configurable search tools via the [Model Context Protocol (MCP)](https://modelcontextprotocol.io), so AI agents like Claude Code can search your docs and code semantically.
+Agentic docs retrieval for AI agents — semantic search and filesystem exploration over your documentation and code via MCP. One config file, one command, works with any AI coding agent.
 
 ## Quick Start
 
-1. **Clone and configure:**
-   ```bash
-   git clone https://github.com/CopilotKit/pathfinder.git
-   cd pathfinder
-   cp pathfinder.example.yaml pathfinder.yaml  # edit for your project
-   cp .env.example .env                        # add your OPENAI_API_KEY
-   ```
-
-2. **Start the server:**
-   ```bash
-   docker compose up
-   ```
-
-3. **Seed the index:**
-   ```bash
-   docker compose exec app npx tsx scripts/seed-index.ts
-   ```
-
-4. **Connect your AI agent:**
-   ```json
-   {
-     "mcpServers": {
-       "my-docs": { "url": "http://localhost:3001/mcp" }
-     }
-   }
-   ```
-
-## Configuration
-
-All configuration lives in `pathfinder.yaml`. See [pathfinder.example.yaml](pathfinder.example.yaml) for a minimal starting point.
-
-### Sources
-
-Each source defines what to index:
-
-```yaml
-sources:
-  - name: docs
-    type: markdown        # Built-in: markdown, code, raw-text
-    repo: https://github.com/your-org/your-repo.git
-    path: docs/
-    base_url: https://docs.your-project.com/
-    url_derivation:
-      strip_prefix: "docs/"
-      strip_suffix: ".md"
-    file_patterns: ["**/*.md"]
-    chunk:
-      target_tokens: 600
-      overlap_tokens: 50
-
-  - name: code
-    type: code
-    repo: https://github.com/your-org/your-repo.git
-    path: "."
-    file_patterns: ["**/*.ts", "**/*.py"]
-    exclude_patterns: ["**/test/**", "**/*.test.*"]
-    chunk:
-      target_lines: 80
-      overlap_lines: 10
-```
-
-### Tools
-
-Each tool maps to a source and defines the MCP tool interface:
-
-```yaml
-tools:
-  - name: search-docs
-    description: "Search documentation for relevant information."
-    source: docs
-    default_limit: 5
-    max_limit: 20
-    result_format: docs
-```
-
-### Collect Tools
-
-Collect tools let agents write structured data back to the server. Unlike search tools, they don't query anything — they validate the agent's input against a YAML-defined schema and store it as JSON in the database. Use them to gather signal from agents without writing any code.
-
-The first built-in use case is search feedback: agents report whether search results were helpful, what they tried, and what went wrong. This surfaces broken or misleading documentation quickly. But collect tools are generic — you can define any schema for any use case (e.g., broken link reporting, feature requests, error logging).
-
-```yaml
-tools:
-  - name: submit-feedback
-    type: collect
-    description: "Submit feedback on whether search results were helpful."
-    response: "Feedback recorded. Thank you."
-    schema:
-      tool_name:
-        type: string
-        description: "Which search tool was used"
-        required: true
-      rating:
-        type: enum
-        values: ["helpful", "not_helpful"]
-        description: "Whether the results were helpful"
-        required: true
-      comment:
-        type: string
-        description: "What worked or didn't work"
-        required: true
-```
-
-Each field in `schema` supports `type` (`string`, `number`, or `enum`), an optional `description` (shown to the agent), `required` (defaults to false), and `values` (required for `enum` fields). The validated input is written as JSONB to the `collected_data` table along with the tool name and a timestamp.
-
-### Bash Tool Options
-
-Bash tools expose source files as a read-only virtual filesystem that agents can explore with standard commands (`find`, `grep`, `cat`, `ls`, `head`). Several options control behavior:
-
-```yaml
-tools:
-  - name: explore-docs
-    type: bash
-    description: "Explore documentation files"
-    sources: [docs]
-    bash:
-      session_state: true       # Persistent CWD across commands (default: false)
-      grep_strategy: hybrid     # memory | vector | hybrid — enables qmd semantic search (default: memory, no qmd)
-      virtual_files: true       # Auto-generate INDEX.md, SEARCH_TIPS.md (default: false)
-```
-
-- **session_state**: When enabled, `cd` persists across commands within a session. Agents can run `cd /docs` in one tool call and then `ls` or `cat file.md` in the next without repeating the path.
-- **grep_strategy**: Controls whether the `qmd` semantic search command is available. `memory` uses pure in-memory regex only (no `qmd`). `vector` or `hybrid` enable the `qmd` command, which performs semantic search via embeddings plus text `ILIKE`. The `vector` and `hybrid` modes require an `embedding` config block.
-- **virtual_files**: Auto-generates `/INDEX.md` (file listing with descriptions) and `/SEARCH_TIPS.md` (usage guidance) at the root of the virtual filesystem.
-
-Agents can also run the `related` command inside bash tools to find semantically similar files across all mounted sources:
-
 ```bash
-related /docs/concepts/coagents.mdx
+npx @copilotkit/pathfinder init
+npx @copilotkit/pathfinder serve
 ```
 
-This returns a ranked list of files from any source that are semantically related to the given file, useful for discovering cross-references between documentation and code.
-
-When `grep_strategy` is set to `vector` or `hybrid`, agents can use the `qmd` command for semantic search:
+Or with Docker:
 
 ```bash
-qmd "how do I configure authentication"
+docker pull ghcr.io/copilotkit/pathfinder
+docker run -v ./pathfinder.yaml:/app/pathfinder.yaml \
+  -v ./docs:/app/docs -p 3001:3001 \
+  ghcr.io/copilotkit/pathfinder
 ```
 
-This performs a 2-pass search (semantic embeddings + text ILIKE) with dedup and filtering, and returns file:line:content results. Standard `grep` is never intercepted — it always works with standard flags as agents expect.
+Then connect your AI agent:
 
-**Note:** The virtual filesystem is read-only and shared across all MCP sessions for a given tool. Content refreshes on webhook or server restart.
-
-### Built-in Chunker Types
+```json
+{
+  "mcpServers": {
+    "my-docs": { "url": "http://localhost:3001/mcp" }
+  }
+}
+```
 
-| Type | Best For | Splits On |
-|------|----------|-----------|
-| `markdown` | .md, .mdx files | Headings (h2->h3->paragraph->line), preserves code blocks |
-| `code` | Source code files | Blank line boundaries, respects block comments/strings |
-| `raw-text` | Plain text, logs | Paragraph boundaries (double newline) |
+## Try It — Pathfinder on Its Own Docs
 
-### Embedding
+This documentation is indexed by a live Pathfinder instance. Connect your agent to try it:
 
-```yaml
-embedding:
-  provider: openai
-  model: text-embedding-3-small
-  dimensions: 1536
+```bash
+# Claude Code
+claude mcp add pathfinder-docs --transport http https://mcp.pathfinder.copilotkit.dev/mcp
 ```
 
-### Indexing
-
-```yaml
-indexing:
-  auto_reindex: true         # Nightly full reindex
-  reindex_hour_utc: 3        # 3 AM UTC
-  stale_threshold_hours: 24
+```json
+// Claude Desktop / Cursor / any MCP client
+{
+  "mcpServers": {
+    "pathfinder-docs": {
+      "url": "https://mcp.pathfinder.copilotkit.dev/mcp"
+    }
+  }
+}
 ```
 
-## Deploying to Production with Docker
-
-The simplest way to run in production:
-
-1. **Configure:**
-   ```bash
-   cp pathfinder.example.yaml pathfinder.yaml  # edit for your project
-   ```
-
-2. **Set environment variables** in `.env`:
-   ```
-   OPENAI_API_KEY=sk-...
-   POSTGRES_PASSWORD=your-secure-password
-   GITHUB_WEBHOOK_SECRET=your-webhook-secret
-   ```
-
-3. **Deploy:**
-   ```bash
-   docker compose -f docker-compose.prod.yaml up -d
-   ```
-
-4. **Verify:**
-   ```bash
-   curl http://localhost:3001/health | python3 -m json.tool
-   ```
+## What It Does
 
-The health endpoint shows index status, chunk counts, and source states.
+Pathfinder indexes your GitHub repos — docs (Markdown, MDX, HTML) and source code — into a PostgreSQL vector database using OpenAI embeddings. It serves configurable search and filesystem exploration tools via [MCP](https://modelcontextprotocol.io), so AI agents can search your docs semantically and browse files with bash commands.
 
-The server automatically indexes on first boot and runs a nightly reindex at the configured hour.
+| Tool Type | What It Does | Example |
+|-----------|-------------|---------|
+| **Search** | Semantic search over indexed content | `search-docs("how to authenticate")` |
+| **Bash** | Virtual filesystem with find, grep, cat, ls | `explore-docs("cat /docs/quickstart.mdx")` |
+| **Collect** | Structured data collection from agents | `submit-feedback(rating: "helpful")` |
 
-### GitHub Webhooks (optional)
+## Features
 
-For real-time re-indexing on push:
+- **[Semantic Search](https://pathfinder.copilotkit.dev/search)** — pgvector RAG with configurable chunk sizes, overlap, and score thresholds
+- **[Filesystem Exploration](https://pathfinder.copilotkit.dev/search)** — QuickJS WASM sandbox with session state, `qmd` semantic grep, `related` files
+- **[4 Source Types](https://pathfinder.copilotkit.dev/config)** — Markdown, code, raw-text, HTML — with pluggable chunker registry
+- **[Config-Driven](https://pathfinder.copilotkit.dev/config)** — Everything in one `pathfinder.yaml`: sources, tools, embedding, indexing, webhooks
+- **[Client Setup](https://pathfinder.copilotkit.dev/clients)** — Claude Desktop, Claude Code, Cursor, Codex, VS Code, any Streamable HTTP client
+- **[Docker + Railway](https://pathfinder.copilotkit.dev/deploy)** — Container image, docker-compose, Railway one-click
+- **[Auto-Generated Endpoints](https://pathfinder.copilotkit.dev/usage)** — `/llms.txt`, `/llms-full.txt`, `/.well-known/skills/default/skill.md`
+- **[Webhook Reindexing](https://pathfinder.copilotkit.dev/deploy)** — GitHub push triggers incremental reindex
+- **[IP Rate Limiting](https://pathfinder.copilotkit.dev/config)** — Per-IP session caps and configurable TTL
 
-1. Add webhook config to `pathfinder.yaml`:
-   ```yaml
-   webhook:
-     repo_sources:
-       "your-org/your-repo": [docs, code]
-     path_triggers:
-       docs: ["docs/"]
-       code: []
-   ```
-
-2. Configure the webhook on GitHub:
-   - URL: `https://your-server/webhooks/github`
-   - Secret: same as `GITHUB_WEBHOOK_SECRET`
-   - Events: Just `push`
-
-## Deploying to Railway
-
-1. **Run setup:**
-   ```bash
-   ./scripts/setup.sh       # install deps, build Docker images
-   ./scripts/deploy.sh      # create Railway project, set vars, deploy
-   ```
-
-2. **Set custom domain** in Railway dashboard
-
-3. **Configure webhooks:**
-   ```bash
-   ./scripts/setup-webhooks.sh
-   ```
-
-See [OPERATIONS.md](OPERATIONS.md) for the full operations runbook.
-
-## Development
+## CLI
 
 ```bash
-# Local setup
-./scripts/setup.sh
+# Scaffold config
+npx @copilotkit/pathfinder init
 
-# Start dev server (hot reload)
-docker compose up
-
-# Seed index
-docker compose exec app npx tsx scripts/seed-index.ts
-
-# Run unit tests
-npm test
+# Start server (uses PGlite if no DATABASE_URL)
+npx @copilotkit/pathfinder serve
 
-# Test search
-docker compose exec app npx tsx scripts/test-search.ts "your query"
+# Docker with Postgres
+docker compose up
+```
 
-# Run integration tests
-npx tsx scripts/integration-test.ts
+## Switching from Mintlify?
 
-# Run path filter tests
-npx tsx scripts/test-path-filter.ts
-```
+Step-by-step migration guide: **[Migrate from Mintlify](https://pathfinder.copilotkit.dev/migrate-from-mintlify)**
 
-## Environment Variables
+## Documentation
 
-| Variable | Required | Description |
-|----------|----------|-------------|
-| `OPENAI_API_KEY` | Yes | OpenAI API key for embeddings |
-| `DATABASE_URL` | Yes | PostgreSQL connection string |
-| `GITHUB_WEBHOOK_SECRET` | No | HMAC secret for webhook verification |
-| `GITHUB_TOKEN` | No | GitHub token for private repos |
-| `PATHFINDER_CONFIG` | No | Path to config file (default: `./pathfinder.yaml`) |
-| `PORT` | No | Server port (default: `3001`) |
+**[https://pathfinder.copilotkit.dev](https://pathfinder.copilotkit.dev)**
 
 ## License
 
-MIT — see [LICENSE](LICENSE)
+MIT

package/dist/cli.d.ts

@@ -0,0 +1,3 @@
+#!/usr/bin/env node
+export {};
+//# sourceMappingURL=cli.d.ts.map

package/dist/cli.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"cli.d.ts","sourceRoot":"","sources":["../src/cli.ts"],"names":[],"mappings":""}

package/dist/cli.js

@@ -0,0 +1,60 @@
+#!/usr/bin/env node
+import { Command } from "commander";
+import fs from "fs";
+import path from "path";
+import { fileURLToPath } from "url";
+const __dirname = path.dirname(fileURLToPath(import.meta.url));
+const program = new Command();
+program
+    .name("pathfinder")
+    .description("Agentic docs retrieval for AI agents")
+    .version("1.1.0");
+program
+    .command("init")
+    .description("Scaffold a new Pathfinder project in the current directory")
+    .action(async () => {
+    const cwd = process.cwd();
+    const yamlDest = path.join(cwd, "pathfinder.yaml");
+    if (fs.existsSync(yamlDest)) {
+        console.log("pathfinder.yaml already exists, skipping.");
+    }
+    else {
+        const templatePath = path.join(__dirname, "..", "pathfinder.example.yaml");
+        if (!fs.existsSync(templatePath)) {
+            console.error("Could not find pathfinder.example.yaml template.");
+            process.exit(1);
+        }
+        fs.copyFileSync(templatePath, yamlDest);
+        console.log("Created pathfinder.yaml");
+    }
+    const envDest = path.join(cwd, ".env");
+    if (fs.existsSync(envDest)) {
+        console.log(".env already exists, skipping.");
+    }
+    else {
+        const envTemplatePath = path.join(__dirname, "..", ".env.example");
+        if (fs.existsSync(envTemplatePath)) {
+            fs.copyFileSync(envTemplatePath, envDest);
+            console.log("Created .env from template");
+        }
+        else {
+            console.log("No .env.example found, skipping .env creation.");
+        }
+    }
+    console.log("\nEdit pathfinder.yaml to configure your docs sources.");
+    console.log("Then run: pathfinder serve");
+});
+program
+    .command("serve")
+    .description("Start the Pathfinder MCP server")
+    .option("-p, --port <port>", "Port to listen on", parseInt)
+    .option("-c, --config <path>", "Path to pathfinder.yaml")
+    .action(async (opts) => {
+    const { startServer } = await import("./server.js");
+    await startServer({
+        port: opts.port,
+        configPath: opts.config,
+    });
+});
+program.parse();
+//# sourceMappingURL=cli.js.map

package/dist/cli.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"cli.js","sourceRoot":"","sources":["../src/cli.ts"],"names":[],"mappings":";AACA,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAC;AACpC,OAAO,EAAE,MAAM,IAAI,CAAC;AACpB,OAAO,IAAI,MAAM,MAAM,CAAC;AACxB,OAAO,EAAE,aAAa,EAAE,MAAM,KAAK,CAAC;AAEpC,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;AAE/D,MAAM,OAAO,GAAG,IAAI,OAAO,EAAE,CAAC;AAE9B,OAAO;KACF,IAAI,CAAC,YAAY,CAAC;KAClB,WAAW,CAAC,sCAAsC,CAAC;KACnD,OAAO,CAAC,OAAO,CAAC,CAAC;AAEtB,OAAO;KACF,OAAO,CAAC,MAAM,CAAC;KACf,WAAW,CAAC,4DAA4D,CAAC;KACzE,MAAM,CAAC,KAAK,IAAI,EAAE;IACf,MAAM,GAAG,GAAG,OAAO,CAAC,GAAG,EAAE,CAAC;IAE1B,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,GAAG,EAAE,iBAAiB,CAAC,CAAC;IACnD,IAAI,EAAE,CAAC,UAAU,CAAC,QAAQ,CAAC,EAAE,CAAC;QAC1B,OAAO,CAAC,GAAG,CAAC,2CAA2C,CAAC,CAAC;IAC7D,CAAC;SAAM,CAAC;QACJ,MAAM,YAAY,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,EAAE,yBAAyB,CAAC,CAAC;QAC3E,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,YAAY,CAAC,EAAE,CAAC;YAC/B,OAAO,CAAC,KAAK,CAAC,kDAAkD,CAAC,CAAC;YAClE,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACpB,CAAC;QACD,EAAE,CAAC,YAAY,CAAC,YAAY,EAAE,QAAQ,CAAC,CAAC;QACxC,OAAO,CAAC,GAAG,CAAC,yBAAyB,CAAC,CAAC;IAC3C,CAAC;IAED,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,GAAG,EAAE,MAAM,CAAC,CAAC;IACvC,IAAI,EAAE,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;QACzB,OAAO,CAAC,GAAG,CAAC,gCAAgC,CAAC,CAAC;IAClD,CAAC;SAAM,CAAC;QACJ,MAAM,eAAe,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,EAAE,cAAc,CAAC,CAAC;QACnE,IAAI,EAAE,CAAC,UAAU,CAAC,eAAe,CAAC,EAAE,CAAC;YACjC,EAAE,CAAC,YAAY,CAAC,eAAe,EAAE,OAAO,CAAC,CAAC;YAC1C,OAAO,CAAC,GAAG,CAAC,4BAA4B,CAAC,CAAC;QAC9C,CAAC;aAAM,CAAC;YACJ,OAAO,CAAC,GAAG,CAAC,gDAAgD,CAAC,CAAC;QAClE,CAAC;IACL,CAAC;IAED,OAAO,CAAC,GAAG,CAAC,wDAAwD,CAAC,CAAC;IACtE,OAAO,CAAC,GAAG,CAAC,4BAA4B,CAAC,CAAC;AAC9C,CAAC,CAAC,CAAC;AAEP,OAAO;KACF,OAAO,CAAC,OAAO,CAAC;KAChB,WAAW,CAAC,iCAAiC,CAAC;KAC9C,MAAM,CAAC,mBAAmB,EAAE,mBAAmB,EAAE,QAAQ,CAAC;KAC1D,MAAM,CAAC,qBAAqB,EAAE,yBAAyB,CAAC;KACxD,MAAM,CAAC,KAAK,EAAE,IAAI,EAAE,EAAE;IACnB,MAAM,EAAE,WAAW,EAAE,GAAG,MAAM,MAAM,CAAC,aAAa,CAAC,CAAC;IACpD,MAAM,WAAW,CAAC;QACd,IAAI,EAAE,IAAI,CAAC,IAAI;QACf,UAAU,EAAE,IAAI,CAAC,MAAM;KAC1B,CAAC,CAAC;AACP,CAAC,CAAC,CAAC;AAEP,OAAO,CAAC,KAAK,EAAE,CAAC"}

package/dist/db/client.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"client.d.ts","sourceRoot":"","sources":["../../src/db/client.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,IAAI,CAAC;AAOpB;;;;;GAKG;AACH,wBAAgB,OAAO,IAAI,EAAE,CAAC,IAAI,CAsBjC;AAwDD;;;;;;;GAOG;AACH,wBAAsB,gBAAgB,IAAI,OAAO,CAAC,IAAI,CAAC,CAyCtD;AAED;;GAEG;AACH,wBAAsB,SAAS,IAAI,OAAO,CAAC,IAAI,CAAC,CAM/C"}
+{"version":3,"file":"client.d.ts","sourceRoot":"","sources":["../../src/db/client.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,IAAI,CAAC;AAOpB;;;;;GAKG;AACH,wBAAgB,OAAO,IAAI,EAAE,CAAC,IAAI,CAsBjC;AAyDD;;;;;;;GAOG;AACH,wBAAsB,gBAAgB,IAAI,OAAO,CAAC,IAAI,CAAC,CA0CtD;AAED;;GAEG;AACH,wBAAsB,SAAS,IAAI,OAAO,CAAC,IAAI,CAAC,CAM/C"}

package/dist/db/client.js

@@ -1,6 +1,6 @@
 import pg from "pg";
 import pgvector from "pgvector/pg";
-import { generateSchema, generateMigration } from "./schema.js";
+import { generateSchema, generateMigration, generatePostSchemaMigration } from "./schema.js";
 import { getConfig, getServerConfig } from "../config.js";
 let pool = null;
 /**
@@ -48,6 +48,7 @@ async function initializePGlite() {
     try {
         await db.exec(generateMigration());
         await db.exec(generateSchema(dimensions));
+        await db.exec(generatePostSchemaMigration());
         await db.exec('COMMIT');
     }
     catch (err) {
@@ -111,6 +112,7 @@ export async function initializeSchema() {
         await migrationClient.query('BEGIN');
         await migrationClient.query(generateMigration());
         await migrationClient.query(generateSchema(dimensions));
+        await migrationClient.query(generatePostSchemaMigration());
         await migrationClient.query('COMMIT');
     }
     catch (err) {

package/dist/db/client.js.map

@@ -1 +1 @@
-{"version":3,"file":"client.js","sourceRoot":"","sources":["../../src/db/client.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,IAAI,CAAC;AACpB,OAAO,QAAQ,MAAM,aAAa,CAAC;AACnC,OAAO,EAAE,cAAc,EAAE,iBAAiB,EAAE,MAAM,aAAa,CAAC;AAChE,OAAO,EAAE,SAAS,EAAE,eAAe,EAAE,MAAM,cAAc,CAAC;AAE1D,IAAI,IAAI,GAAmB,IAAI,CAAC;AAEhC;;;;;GAKG;AACH,MAAM,UAAU,OAAO;IACnB,IAAI,IAAI;QAAE,OAAO,IAAI,CAAC;IAEtB,MAAM,WAAW,GAAG,SAAS,EAAE,CAAC,WAAW,CAAC;IAE5C,IAAI,CAAC,WAAW,EAAE,CAAC;QACf,MAAM,IAAI,KAAK,CACX,uDAAuD,CAC1D,CAAC;IACN,CAAC;IAED,IAAI,WAAW,CAAC,WAAW,CAAC,EAAE,CAAC;QAC3B,MAAM,IAAI,KAAK,CACX,6DAA6D,CAChE,CAAC;IACN,CAAC;IAED,IAAI,GAAG,IAAI,EAAE,CAAC,IAAI,CAAC;QACf,gBAAgB,EAAE,WAAW;KAChC,CAAC,CAAC;IAEH,OAAO,IAAI,CAAC;AAChB,CAAC;AAED,SAAS,WAAW,CAAC,GAAuB;IACxC,OAAO,CAAC,CAAC,GAAG,IAAI,GAAG,CAAC,UAAU,CAAC,WAAW,CAAC,CAAC;AAChD,CAAC;AAED,SAAS,kBAAkB,CAAC,GAAW;IACnC,OAAO,GAAG,CAAC,OAAO,CAAC,cAAc,EAAE,EAAE,CAAC,CAAC;AAC3C,CAAC;AAED,KAAK,UAAU,gBAAgB;IAC3B,MAAM,WAAW,GAAG,SAAS,EAAE,CAAC,WAAW,CAAC;IAC5C,IAAI,CAAC,WAAW,EAAE,CAAC;QACf,MAAM,IAAI,KAAK,CAAC,oDAAoD,CAAC,CAAC;IAC1E,CAAC;IACD,MAAM,OAAO,GAAG,kBAAkB,CAAC,WAAW,CAAC,CAAC;IAChD,MAAM,UAAU,GAAG,eAAe,EAAE,CAAC,SAAS,EAAE,UAAU,CAAC;IAC3D,IAAI,CAAC,UAAU;QAAE,MAAM,IAAI,KAAK,CAAC,qEAAqE,CAAC,CAAC;IAExG,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,MAAM,CAAC,sBAAsB,CAAC,CAAC;IACxD,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,MAAM,CAAC,6BAA6B,CAAC,CAAC;IAE/D,MAAM,EAAE,GAAG,IAAI,MAAM,CAAC,EAAE,OAAO,EAAE,UAAU,EAAE,EAAE,MAAM,EAAE,EAAE,CAAC,CAAC;IAC3D,MAAM,EAAE,CAAC,SAAS,CAAC;IAEnB,6DAA6D;IAC7D,MAAM,EAAE,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IACvB,IAAI,CAAC;QACD,MAAM,EAAE,CAAC,IAAI,CAAC,iBAAiB,EAAE,CAAC,CAAC;QACnC,MAAM,EAAE,CAAC,IAAI,CAAC,cAAc,CAAC,UAAU,CAAC,CAAC,CAAC;QAC1C,MAAM,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;IAC5B,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACX,IAAI,CAAC;YACD,MAAM,EAAE,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;QAC9B,CAAC;QAAC,MAAM,CAAC;YACL,kDAAkD;QACtD,CAAC;QACD,MAAM,GAAG,CAAC;IACd,CAAC;IAED,8CAA8C;IAC9C,wFAAwF;IACxF,iFAAiF;IACjF,iFAAiF;IACjF,MAAM,OAAO,GAAG;QACZ,KAAK,EAAE,CAAC,IAAY,EAAE,MAAkB,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,CAAC,IAAI,EAAE,MAAM,CAAC;QACnE,OAAO,EAAE,KAAK,IAAI,EAAE,CAAC,CAAC;YAClB,KAAK,EAAE,CAAC,IAAY,EAAE,MAAkB,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,CAAC,IAAI,EAAE,MAAM,CAAC;YACnE,OAAO,EAAE,GAAG,EAAE,GAAE,CAAC;SACpB,CAAC;QACF,GAAG,EAAE,KAAK,IAAI,EAAE,CAAC,EAAE,CAAC,KAAK,EAAE;KAC9B,CAAC;IAEF,IAAI,GAAG,OAA6B,CAAC;AACzC,CAAC;AAED;;;;;;;GAOG;AACH,MAAM,CAAC,KAAK,UAAU,gBAAgB;IAClC,MAAM,WAAW,GAAG,SAAS,EAAE,CAAC,WAAW,CAAC;IAE5C,IAAI,CAAC,WAAW,EAAE,CAAC;QACf,MAAM,IAAI,KAAK,CAAC,6DAA6D,CAAC,CAAC;IACnF,CAAC;IAED,IAAI,WAAW,CAAC,WAAW,CAAC,EAAE,CAAC;QAC3B,MAAM,gBAAgB,EAAE,CAAC;QACzB,OAAO;IACX,CAAC;IAED,MAAM,CAAC,GAAG,OAAO,EAAE,CAAC;IAEpB,MAAM,UAAU,GAAG,eAAe,EAAE,CAAC,SAAS,EAAE,UAAU,CAAC;IAC3D,IAAI,CAAC,UAAU;QAAE,MAAM,IAAI,KAAK,CAAC,qEAAqE,CAAC,CAAC;IAExG,8DAA8D;IAC9D,MAAM,WAAW,GAAG,MAAM,CAAC,CAAC,OAAO,EAAE,CAAC;IACtC,IAAI,CAAC;QACD,iFAAiF;QACjF,yEAAyE;QACzE,MAAM,WAAW,CAAC,KAAK,CAAC,uCAAuC,CAAC,CAAC;QACjE,MAAM,QAAQ,CAAC,YAAY,CAAC,WAAW,CAAC,CAAC;IAC7C,CAAC;YAAS,CAAC;QACP,WAAW,CAAC,OAAO,EAAE,CAAC;IAC1B,CAAC;IAED,6CAA6C;IAC7C,MAAM,eAAe,GAAG,MAAM,CAAC,CAAC,OAAO,EAAE,CAAC;IAC1C,IAAI,CAAC;QACD,MAAM,eAAe,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;QACrC,MAAM,eAAe,CAAC,KAAK,CAAC,iBAAiB,EAAE,CAAC,CAAC;QACjD,MAAM,eAAe,CAAC,KAAK,CAAC,cAAc,CAAC,UAAU,CAAC,CAAC,CAAC;QACxD,MAAM,eAAe,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;IAC1C,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACX,MAAM,eAAe,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC;QACxC,MAAM,GAAG,CAAC;IACd,CAAC;YAAS,CAAC;QACP,eAAe,CAAC,OAAO,EAAE,CAAC;IAC9B,CAAC;AACL,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,SAAS;IAC3B,IAAI,IAAI,EAAE,CAAC;QACP,MAAM,CAAC,GAAG,IAAI,CAAC;QACf,IAAI,GAAG,IAAI,CAAC;QACZ,MAAM,CAAC,CAAC,GAAG,EAAE,CAAC;IAClB,CAAC;AACL,CAAC"}
+{"version":3,"file":"client.js","sourceRoot":"","sources":["../../src/db/client.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,IAAI,CAAC;AACpB,OAAO,QAAQ,MAAM,aAAa,CAAC;AACnC,OAAO,EAAE,cAAc,EAAE,iBAAiB,EAAE,2BAA2B,EAAE,MAAM,aAAa,CAAC;AAC7F,OAAO,EAAE,SAAS,EAAE,eAAe,EAAE,MAAM,cAAc,CAAC;AAE1D,IAAI,IAAI,GAAmB,IAAI,CAAC;AAEhC;;;;;GAKG;AACH,MAAM,UAAU,OAAO;IACnB,IAAI,IAAI;QAAE,OAAO,IAAI,CAAC;IAEtB,MAAM,WAAW,GAAG,SAAS,EAAE,CAAC,WAAW,CAAC;IAE5C,IAAI,CAAC,WAAW,EAAE,CAAC;QACf,MAAM,IAAI,KAAK,CACX,uDAAuD,CAC1D,CAAC;IACN,CAAC;IAED,IAAI,WAAW,CAAC,WAAW,CAAC,EAAE,CAAC;QAC3B,MAAM,IAAI,KAAK,CACX,6DAA6D,CAChE,CAAC;IACN,CAAC;IAED,IAAI,GAAG,IAAI,EAAE,CAAC,IAAI,CAAC;QACf,gBAAgB,EAAE,WAAW;KAChC,CAAC,CAAC;IAEH,OAAO,IAAI,CAAC;AAChB,CAAC;AAED,SAAS,WAAW,CAAC,GAAuB;IACxC,OAAO,CAAC,CAAC,GAAG,IAAI,GAAG,CAAC,UAAU,CAAC,WAAW,CAAC,CAAC;AAChD,CAAC;AAED,SAAS,kBAAkB,CAAC,GAAW;IACnC,OAAO,GAAG,CAAC,OAAO,CAAC,cAAc,EAAE,EAAE,CAAC,CAAC;AAC3C,CAAC;AAED,KAAK,UAAU,gBAAgB;IAC3B,MAAM,WAAW,GAAG,SAAS,EAAE,CAAC,WAAW,CAAC;IAC5C,IAAI,CAAC,WAAW,EAAE,CAAC;QACf,MAAM,IAAI,KAAK,CAAC,oDAAoD,CAAC,CAAC;IAC1E,CAAC;IACD,MAAM,OAAO,GAAG,kBAAkB,CAAC,WAAW,CAAC,CAAC;IAChD,MAAM,UAAU,GAAG,eAAe,EAAE,CAAC,SAAS,EAAE,UAAU,CAAC;IAC3D,IAAI,CAAC,UAAU;QAAE,MAAM,IAAI,KAAK,CAAC,qEAAqE,CAAC,CAAC;IAExG,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,MAAM,CAAC,sBAAsB,CAAC,CAAC;IACxD,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,MAAM,CAAC,6BAA6B,CAAC,CAAC;IAE/D,MAAM,EAAE,GAAG,IAAI,MAAM,CAAC,EAAE,OAAO,EAAE,UAAU,EAAE,EAAE,MAAM,EAAE,EAAE,CAAC,CAAC;IAC3D,MAAM,EAAE,CAAC,SAAS,CAAC;IAEnB,6DAA6D;IAC7D,MAAM,EAAE,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IACvB,IAAI,CAAC;QACD,MAAM,EAAE,CAAC,IAAI,CAAC,iBAAiB,EAAE,CAAC,CAAC;QACnC,MAAM,EAAE,CAAC,IAAI,CAAC,cAAc,CAAC,UAAU,CAAC,CAAC,CAAC;QAC1C,MAAM,EAAE,CAAC,IAAI,CAAC,2BAA2B,EAAE,CAAC,CAAC;QAC7C,MAAM,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;IAC5B,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACX,IAAI,CAAC;YACD,MAAM,EAAE,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;QAC9B,CAAC;QAAC,MAAM,CAAC;YACL,kDAAkD;QACtD,CAAC;QACD,MAAM,GAAG,CAAC;IACd,CAAC;IAED,8CAA8C;IAC9C,wFAAwF;IACxF,iFAAiF;IACjF,iFAAiF;IACjF,MAAM,OAAO,GAAG;QACZ,KAAK,EAAE,CAAC,IAAY,EAAE,MAAkB,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,CAAC,IAAI,EAAE,MAAM,CAAC;QACnE,OAAO,EAAE,KAAK,IAAI,EAAE,CAAC,CAAC;YAClB,KAAK,EAAE,CAAC,IAAY,EAAE,MAAkB,EAAE,EAAE,CAAC,EAAE,CAAC,KAAK,CAAC,IAAI,EAAE,MAAM,CAAC;YACnE,OAAO,EAAE,GAAG,EAAE,GAAE,CAAC;SACpB,CAAC;QACF,GAAG,EAAE,KAAK,IAAI,EAAE,CAAC,EAAE,CAAC,KAAK,EAAE;KAC9B,CAAC;IAEF,IAAI,GAAG,OAA6B,CAAC;AACzC,CAAC;AAED;;;;;;;GAOG;AACH,MAAM,CAAC,KAAK,UAAU,gBAAgB;IAClC,MAAM,WAAW,GAAG,SAAS,EAAE,CAAC,WAAW,CAAC;IAE5C,IAAI,CAAC,WAAW,EAAE,CAAC;QACf,MAAM,IAAI,KAAK,CAAC,6DAA6D,CAAC,CAAC;IACnF,CAAC;IAED,IAAI,WAAW,CAAC,WAAW,CAAC,EAAE,CAAC;QAC3B,MAAM,gBAAgB,EAAE,CAAC;QACzB,OAAO;IACX,CAAC;IAED,MAAM,CAAC,GAAG,OAAO,EAAE,CAAC;IAEpB,MAAM,UAAU,GAAG,eAAe,EAAE,CAAC,SAAS,EAAE,UAAU,CAAC;IAC3D,IAAI,CAAC,UAAU;QAAE,MAAM,IAAI,KAAK,CAAC,qEAAqE,CAAC,CAAC;IAExG,8DAA8D;IAC9D,MAAM,WAAW,GAAG,MAAM,CAAC,CAAC,OAAO,EAAE,CAAC;IACtC,IAAI,CAAC;QACD,iFAAiF;QACjF,yEAAyE;QACzE,MAAM,WAAW,CAAC,KAAK,CAAC,uCAAuC,CAAC,CAAC;QACjE,MAAM,QAAQ,CAAC,YAAY,CAAC,WAAW,CAAC,CAAC;IAC7C,CAAC;YAAS,CAAC;QACP,WAAW,CAAC,OAAO,EAAE,CAAC;IAC1B,CAAC;IAED,6CAA6C;IAC7C,MAAM,eAAe,GAAG,MAAM,CAAC,CAAC,OAAO,EAAE,CAAC;IAC1C,IAAI,CAAC;QACD,MAAM,eAAe,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;QACrC,MAAM,eAAe,CAAC,KAAK,CAAC,iBAAiB,EAAE,CAAC,CAAC;QACjD,MAAM,eAAe,CAAC,KAAK,CAAC,cAAc,CAAC,UAAU,CAAC,CAAC,CAAC;QACxD,MAAM,eAAe,CAAC,KAAK,CAAC,2BAA2B,EAAE,CAAC,CAAC;QAC3D,MAAM,eAAe,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;IAC1C,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACX,MAAM,eAAe,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC;QACxC,MAAM,GAAG,CAAC;IACd,CAAC;YAAS,CAAC;QACP,eAAe,CAAC,OAAO,EAAE,CAAC;IAC9B,CAAC;AACL,CAAC;AAED;;GAEG;AACH,MAAM,CAAC,KAAK,UAAU,SAAS;IAC3B,IAAI,IAAI,EAAE,CAAC;QACP,MAAM,CAAC,GAAG,IAAI,CAAC;QACf,IAAI,GAAG,IAAI,CAAC;QACZ,MAAM,CAAC,CAAC,GAAG,EAAE,CAAC;IAClB,CAAC;AACL,CAAC"}

package/dist/db/queries.d.ts

@@ -1,10 +1,10 @@
 import type { Chunk, ChunkResult, IndexState } from "../types.js";
 /**
  * Cosine similarity search on the unified chunks table.
- * Optionally filtered by source_name. Returns results ordered by similarity
- * (highest first).
+ * Optionally filtered by source_name and/or version. Returns results ordered
+ * by similarity (highest first).
  */
-export declare function searchChunks(embedding: number[], limit: number, sourceName?: string): Promise<ChunkResult[]>;
+export declare function searchChunks(embedding: number[], limit: number, sourceName?: string, version?: string): Promise<ChunkResult[]>;
 /**
  * Text search (ILIKE) on the unified chunks table.
  * Optionally filtered by source_name. Returns results ordered by id.
@@ -44,6 +44,17 @@ export interface IndexStats {
     indexedRepos: number;
     indexStates: IndexState[];
 }
+/**
+ * Fetch all chunks (without embeddings) for llms.txt generation.
+ * Ordered by source_name, file_path, chunk_index for deterministic output.
+ */
+export declare function getAllChunksForLlms(): Promise<{
+    source_name: string;
+    file_path: string;
+    title: string | null;
+    content: string;
+    chunk_index: number;
+}[]>;
 /**
  * Get aggregate statistics for the health endpoint.
  */

package/dist/db/queries.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"queries.d.ts","sourceRoot":"","sources":["../../src/db/queries.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EAAE,KAAK,EAAE,WAAW,EAAE,UAAU,EAAe,MAAM,aAAa,CAAC;AAM/E;;;;GAIG;AACH,wBAAsB,YAAY,CAC9B,SAAS,EAAE,MAAM,EAAE,EACnB,KAAK,EAAE,MAAM,EACb,UAAU,CAAC,EAAE,MAAM,GACpB,OAAO,CAAC,WAAW,EAAE,CAAC,CA6DxB;AAED;;;GAGG;AACH,wBAAsB,gBAAgB,CAClC,OAAO,EAAE,MAAM,EACf,KAAK,EAAE,MAAM,EACb,UAAU,CAAC,EAAE,MAAM,GACpB,OAAO,CAAC,WAAW,EAAE,CAAC,CA2BxB;AAMD;;;GAGG;AACH,wBAAsB,YAAY,CAAC,MAAM,EAAE,KAAK,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC,CAuDjE;AAMD;;GAEG;AACH,wBAAsB,kBAAkB,CACpC,UAAU,EAAE,MAAM,EAClB,QAAQ,EAAE,MAAM,GACjB,OAAO,CAAC,IAAI,CAAC,CAMf;AAED;;GAEG;AACH,wBAAsB,oBAAoB,CAAC,UAAU,EAAE,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC,CAG5E;AAMD;;GAEG;AACH,wBAAsB,aAAa,CAC/B,UAAU,EAAE,MAAM,EAClB,SAAS,EAAE,MAAM,GAClB,OAAO,CAAC,UAAU,GAAG,IAAI,CAAC,CAmB5B;AAED;;GAEG;AACH,wBAAsB,gBAAgB,CAAC,KAAK,EAAE,UAAU,GAAG,OAAO,CAAC,IAAI,CAAC,CAqBvE;AAMD;;GAEG;AACH,wBAAsB,mBAAmB,CACrC,QAAQ,EAAE,MAAM,EAChB,IAAI,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC9B,OAAO,CAAC,IAAI,CAAC,CAMf;AAMD,MAAM,WAAW,UAAU;IACvB,WAAW,EAAE,MAAM,CAAC;IACpB,QAAQ,EAAE,KAAK,CAAC;QAAE,WAAW,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IACxD,YAAY,EAAE,MAAM,CAAC;IACrB,WAAW,EAAE,UAAU,EAAE,CAAC;CAC7B;AAED;;GAEG;AACH,wBAAsB,aAAa,IAAI,OAAO,CAAC,UAAU,CAAC,CA8BzD"}
+{"version":3,"file":"queries.d.ts","sourceRoot":"","sources":["../../src/db/queries.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EAAE,KAAK,EAAE,WAAW,EAAE,UAAU,EAAe,MAAM,aAAa,CAAC;AAM/E;;;;GAIG;AACH,wBAAsB,YAAY,CAC9B,SAAS,EAAE,MAAM,EAAE,EACnB,KAAK,EAAE,MAAM,EACb,UAAU,CAAC,EAAE,MAAM,EACnB,OAAO,CAAC,EAAE,MAAM,GACjB,OAAO,CAAC,WAAW,EAAE,CAAC,CAsDxB;AAED;;;GAGG;AACH,wBAAsB,gBAAgB,CAClC,OAAO,EAAE,MAAM,EACf,KAAK,EAAE,MAAM,EACb,UAAU,CAAC,EAAE,MAAM,GACpB,OAAO,CAAC,WAAW,EAAE,CAAC,CA2BxB;AAMD;;;GAGG;AACH,wBAAsB,YAAY,CAAC,MAAM,EAAE,KAAK,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC,CAyDjE;AAMD;;GAEG;AACH,wBAAsB,kBAAkB,CACpC,UAAU,EAAE,MAAM,EAClB,QAAQ,EAAE,MAAM,GACjB,OAAO,CAAC,IAAI,CAAC,CAMf;AAED;;GAEG;AACH,wBAAsB,oBAAoB,CAAC,UAAU,EAAE,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC,CAG5E;AAMD;;GAEG;AACH,wBAAsB,aAAa,CAC/B,UAAU,EAAE,MAAM,EAClB,SAAS,EAAE,MAAM,GAClB,OAAO,CAAC,UAAU,GAAG,IAAI,CAAC,CAmB5B;AAED;;GAEG;AACH,wBAAsB,gBAAgB,CAAC,KAAK,EAAE,UAAU,GAAG,OAAO,CAAC,IAAI,CAAC,CAqBvE;AAMD;;GAEG;AACH,wBAAsB,mBAAmB,CACrC,QAAQ,EAAE,MAAM,EAChB,IAAI,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,GAC9B,OAAO,CAAC,IAAI,CAAC,CAMf;AAMD,MAAM,WAAW,UAAU;IACvB,WAAW,EAAE,MAAM,CAAC;IACpB,QAAQ,EAAE,KAAK,CAAC;QAAE,WAAW,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,CAAC,CAAC;IACxD,YAAY,EAAE,MAAM,CAAC;IACrB,WAAW,EAAE,UAAU,EAAE,CAAC;CAC7B;AAED;;;GAGG;AACH,wBAAsB,mBAAmB,IAAI,OAAO,CAAC;IAAE,WAAW,EAAE,MAAM,CAAC;IAAC,SAAS,EAAE,MAAM,CAAC;IAAC,KAAK,EAAE,MAAM,GAAG,IAAI,CAAC;IAAC,OAAO,EAAE,MAAM,CAAC;IAAC,WAAW,EAAE,MAAM,CAAA;CAAE,EAAE,CAAC,CAM7J;AAED;;GAEG;AACH,wBAAsB,aAAa,IAAI,OAAO,CAAC,UAAU,CAAC,CA8BzD"}

package/dist/db/queries.js

@@ -5,54 +5,44 @@ import { getPool } from "./client.js";
 // ---------------------------------------------------------------------------
 /**
  * Cosine similarity search on the unified chunks table.
- * Optionally filtered by source_name. Returns results ordered by similarity
- * (highest first).
+ * Optionally filtered by source_name and/or version. Returns results ordered
+ * by similarity (highest first).
  */
-export async function searchChunks(embedding, limit, sourceName) {
+export async function searchChunks(embedding, limit, sourceName, version) {
     const pool = getPool();
-    let sql;
-    let params;
+    const conditions = [];
+    const params = [pgvector.toSql(embedding)];
+    let paramIdx = 2;
     if (sourceName) {
-        sql = `
-            SELECT
-                id,
-                source_name,
-                source_url,
-                title,
-                content,
-                repo_url,
-                file_path,
-                start_line,
-                end_line,
-                language,
-                1 - (embedding <=> $1) AS similarity
-            FROM chunks
-            WHERE source_name = $2
-            ORDER BY embedding <=> $1
-            LIMIT $3
-        `;
-        params = [pgvector.toSql(embedding), sourceName, limit];
+        conditions.push(`source_name = $${paramIdx++}`);
+        params.push(sourceName);
     }
-    else {
-        sql = `
-            SELECT
-                id,
-                source_name,
-                source_url,
-                title,
-                content,
-                repo_url,
-                file_path,
-                start_line,
-                end_line,
-                language,
-                1 - (embedding <=> $1) AS similarity
-            FROM chunks
-            ORDER BY embedding <=> $1
-            LIMIT $2
-        `;
-        params = [pgvector.toSql(embedding), limit];
+    if (version) {
+        conditions.push(`version = $${paramIdx++}`);
+        params.push(version);
     }
+    const whereClause = conditions.length > 0
+        ? `WHERE ${conditions.join(' AND ')}`
+        : '';
+    const sql = `
+        SELECT
+            id,
+            source_name,
+            source_url,
+            title,
+            content,
+            repo_url,
+            file_path,
+            start_line,
+            end_line,
+            language,
+            1 - (embedding <=> $1) AS similarity
+        FROM chunks
+        ${whereClause}
+        ORDER BY embedding <=> $1
+        LIMIT $${paramIdx}
+    `;
+    params.push(limit);
     const { rows } = await pool.query(sql, params);
     return rows.map((r) => ({
         id: r.id,
@@ -119,9 +109,9 @@ export async function upsertChunks(chunks) {
             INSERT INTO chunks
                 (source_name, source_url, title, content, embedding, repo_url,
                  file_path, start_line, end_line, language, chunk_index,
-                 metadata, commit_sha, indexed_at)
+                 metadata, commit_sha, version, indexed_at)
             VALUES
-                ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11, $12, $13, NOW())
+                ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11, $12, $13, $14, NOW())
             ON CONFLICT (source_name, file_path, chunk_index) DO UPDATE SET
                 source_url = EXCLUDED.source_url,
                 title      = EXCLUDED.title,
@@ -133,6 +123,7 @@ export async function upsertChunks(chunks) {
                 language   = EXCLUDED.language,
                 metadata   = EXCLUDED.metadata,
                 commit_sha = EXCLUDED.commit_sha,
+                version    = EXCLUDED.version,
                 indexed_at = NOW()
         `;
         for (const chunk of chunks) {
@@ -150,6 +141,7 @@ export async function upsertChunks(chunks) {
                 chunk.chunk_index,
                 JSON.stringify(chunk.metadata ?? {}),
                 chunk.commit_sha ?? null,
+                chunk.version ?? null,
             ]);
         }
         await client.query("COMMIT");
@@ -240,6 +232,15 @@ export async function insertCollectedData(toolName, data) {
     const pool = getPool();
     await pool.query("INSERT INTO collected_data (tool_name, data) VALUES ($1, $2)", [toolName, JSON.stringify(data)]);
 }
+/**
+ * Fetch all chunks (without embeddings) for llms.txt generation.
+ * Ordered by source_name, file_path, chunk_index for deterministic output.
+ */
+export async function getAllChunksForLlms() {
+    const pool = getPool();
+    const result = await pool.query('SELECT source_name, file_path, title, content, chunk_index FROM chunks ORDER BY source_name, file_path, chunk_index');
+    return result.rows;
+}
 /**
  * Get aggregate statistics for the health endpoint.
  */