mcp-local-rag 0.9.0 → 0.10.1

package/README.md

@@ -1,3 +1,7 @@
+<p align="center">
+  <img src="assets/banner.jpg" alt="MCP Local RAG — Search below the surface." width="600" />
+</p>
+
 # MCP Local RAG
 
 [![GitHub stars](https://img.shields.io/github/stars/shinpr/mcp-local-rag?style=social)](https://github.com/shinpr/mcp-local-rag)
@@ -6,7 +10,7 @@
 [![TypeScript](https://img.shields.io/badge/TypeScript-5.0-blue.svg?logo=typescript&logoColor=white)](https://www.typescriptlang.org/)
 [![MCP Registry](https://img.shields.io/badge/MCP-Registry-green.svg)](https://registry.modelcontextprotocol.io/)
 
-Local RAG for developers using MCP.
+Local RAG for developers via MCP or CLI.
 Semantic search with keyword boost for exact technical terms — fully private, zero setup.
 
 ## Features
@@ -24,7 +28,8 @@ Semantic search with keyword boost for exact technical terms — fully private,
   No API keys, no cloud, no data leaving your machine. Works fully offline after the first model download.
 
 - **Zero-friction setup**
-  One `npx` command. No Docker, no Python, no servers to manage. Designed for Cursor, Codex, and Claude Code via MCP.
+  One `npx` command. No Docker, no Python, no servers to manage.
+  Use via MCP, CLI, or both. Optional [Agent Skills](#agent-skills) help AI assistants form better queries and interpret results.
 
 ## Quick Start
 
@@ -73,7 +78,14 @@ Assistant: Based on the documentation, authentication uses OAuth 2.0 with JWT to
           The flow is described in section 3.2...
 ```
 
-That's it. No installation, no Docker, no complex setup.
+**Or use directly as CLI** — no MCP server needed:
+
+```bash
+npx mcp-local-rag ingest ./docs/
+npx mcp-local-rag query "authentication API"
+```
+
+That's it. No Docker, no Python, no server setup.
 
 ## Why This Exists
 
@@ -87,12 +99,17 @@ You want AI to search your documents—technical specs, research papers, interna
 
 **Code search.** Pure semantic search misses exact terms like `useEffect` or `ERR_CONNECTION_REFUSED`. Keyword boost catches both meaning and exact matches.
 
+**Agent reality.** In practice, many AI environments mainly use tool calling. CLI support and Agent Skills make the same workflows available even without full MCP integration.
+
 ## Usage
 
-The server provides 6 MCP tools: ingest file, ingest data, search, list, delete, status
-(`ingest_file`, `ingest_data`, `query_documents`, `list_files`, `delete_file`, `status`).
+mcp-local-rag provides two interfaces: an **MCP server** for AI coding tools and a **CLI** for direct use from the terminal.
+
+### Using with MCP
 
-### Ingesting Documents
+The MCP server provides 6 tools: `ingest_file`, `ingest_data`, `query_documents`, `list_files`, `delete_file`, `status`.
+
+#### Ingesting Documents
 
 ```
 "Ingest the document at /Users/me/docs/api-spec.pdf"
@@ -102,7 +119,7 @@ Supports PDF, DOCX, TXT, and Markdown. The server extracts text, splits it into
 
 Re-ingesting the same file replaces the old version automatically.
 
-### Ingesting HTML Content
+#### Ingesting HTML Content
 
 Use `ingest_data` to ingest HTML content retrieved by your AI assistant (via web fetch, curl, browser tools, etc.):
 
@@ -119,31 +136,7 @@ HTML is automatically cleaned—you get the article content, not the boilerplate
 
 > **Note:** The RAG server itself doesn't fetch web content—your AI assistant retrieves it and passes the HTML to `ingest_data`. This keeps the server fully local while letting you index any content your assistant can access. Please respect website terms of service and copyright when ingesting external content.
 
-### Bulk Ingestion (CLI)
-
-For ingesting multiple files or an entire directory, use the CLI command instead of calling `ingest_file` repeatedly:
-
-```bash
-npx mcp-local-rag ingest --db-path ./lancedb --base-dir ./docs ./docs/
-```
-
-Supports PDF, DOCX, TXT, and Markdown (same as MCP tools). HTML is only supported via `ingest_data`, not CLI.
-
-Key options:
-
-| Option | Default | Description |
-|--------|---------|-------------|
-| `--db-path <path>` | `./lancedb/` | LanceDB database path |
-| `--base-dir <path>` | cwd | Base directory for documents |
-| `--model-name <name>` | `Xenova/all-MiniLM-L6-v2` | Embedding model |
-| `--cache-dir <path>` | `./models/` | Model cache directory |
-| `--max-file-size <n>` | `104857600` | Max file size in bytes |
-
-This processes all supported files recursively and runs optimization once at the end — much faster than per-file ingestion for large batches.
-
-> ⚠️ **CLI options must match your MCP server config.** Especially `--model-name` — using a different embedding model against an existing database produces incompatible vectors, silently degrading search quality.
-
-### Searching Documents
+#### Searching Documents
 
 ```
 "What does the API documentation say about authentication?"
@@ -155,7 +148,7 @@ Search uses semantic similarity with keyword boost. This means `useEffect` finds
 
 Results include text content, source file, document title, and relevance score. The document title provides context for each chunk, helping identify which document a result belongs to. Adjust result count with `limit` (1-20, default 10).
 
-### Managing Files
+#### Managing Files
 
 ```
 "List all files in BASE_DIR and their ingested status"   # See what's indexed
@@ -163,6 +156,58 @@ Results include text content, source file, document title, and relevance score.
 "Show RAG server status"           # Check system health
 ```
 
+### Using as CLI
+
+All MCP tools are also available as CLI commands — no MCP server needed:
+
+```bash
+npx mcp-local-rag ingest ./docs/               # Bulk ingest files
+npx mcp-local-rag query "authentication API"    # Search documents
+npx mcp-local-rag list                          # Show ingestion status
+npx mcp-local-rag status                        # Database stats
+npx mcp-local-rag delete ./docs/old.pdf         # Remove content
+npx mcp-local-rag delete --source "https://..."  # Remove by source URL
+```
+
+`query`, `list`, `status`, and `delete` output JSON to stdout for piping (e.g., `| jq`). `ingest` outputs progress to stderr. Global options (`--db-path`, `--cache-dir`, `--model-name`) go before the subcommand. Run `npx mcp-local-rag --help` for details.
+
+> ⚠️ The CLI does **not** read your MCP client config (`mcp.json`, `config.toml`, etc.). Configure the CLI via flags or environment variables as shown below.
+
+#### Configuration
+
+**CLI flags** — global options go before the subcommand, subcommand options go after:
+
+```bash
+npx mcp-local-rag --db-path ./my-db query "auth" --base-dir ./docs
+```
+
+**Environment variables** — set in your shell:
+
+```bash
+export DB_PATH=./my-db
+export BASE_DIR=./docs
+npx mcp-local-rag query "auth"
+```
+
+**Sharing config between MCP and CLI** — if your MCP client inherits shell environment variables, you can set them in your shell profile (e.g., `~/.zshrc`) so both use the same values. Otherwise, set them explicitly in your MCP config as well.
+
+```bash
+export BASE_DIR=/path/to/your/documents
+export DB_PATH=/path/to/lancedb
+```
+
+Configuration is resolved in this order:
+
+1. CLI flags (highest priority)
+2. Environment variables
+3. Defaults
+
+For the full list of CLI flags, environment variables, and defaults, see [Configuration](#configuration).
+
+For CLI-only setups (no MCP server), install [Agent Skills](#agent-skills) so your AI assistant can form better queries and interpret results consistently.
+
+> ⚠️ **`--model-name` must match your MCP server config.** Using a different embedding model against an existing database produces incompatible vectors, silently degrading search quality.
+
 ## Search Tuning
 
 Adjust these for your use case:
@@ -247,21 +292,22 @@ Before RAG operations, request in natural language:
 Add to your `AGENTS.md`, `CLAUDE.md`, or other agent instruction file:
 ```
 When using query_documents, ingest_file, or ingest_data tools,
-apply the mcp-local-rag skill for optimal query formulation and result interpretation.
+apply the mcp-local-rag skill for better query formulation and result interpretation.
 ```
 
-<details>
-<summary><strong>Configuration</strong></summary>
+## Configuration
 
-### Environment Variables
+### Environment Variables and CLI Flags
 
-| Variable | Default | Description |
-|----------|---------|-------------|
-| `BASE_DIR` | Current directory | Document root directory (security boundary) |
-| `DB_PATH` | `./lancedb/` | Vector database location |
-| `CACHE_DIR` | `./models/` | Model cache directory |
-| `MODEL_NAME` | `Xenova/all-MiniLM-L6-v2` | HuggingFace model ID ([available models](https://huggingface.co/models?library=transformers.js&pipeline_tag=feature-extraction)) |
-| `MAX_FILE_SIZE` | `104857600` (100MB) | Maximum file size in bytes |
+Both MCP and CLI use the same environment variables. The CLI also accepts equivalent flags.
+
+| Environment Variable | CLI Flag | Default | Description |
+|---------------------|----------|---------|-------------|
+| `BASE_DIR` | `--base-dir` | Current directory | Document root directory (security boundary) |
+| `DB_PATH` | `--db-path` | `./lancedb/` | Vector database location |
+| `CACHE_DIR` | `--cache-dir` | `./models/` | Model cache directory |
+| `MODEL_NAME` | `--model-name` | `Xenova/all-MiniLM-L6-v2` | HuggingFace model ID ([available models](https://huggingface.co/models?library=transformers.js&pipeline_tag=feature-extraction)) |
+| `MAX_FILE_SIZE` | `--max-file-size` | `104857600` (100MB) | Maximum file size in bytes |
 
 **Model choice tips:**
 - Multilingual docs → e.g., `onnx-community/embeddinggemma-300m-ONNX` (100+ languages)
@@ -317,8 +363,6 @@ The embedding model (~90MB) downloads on first use. Takes 1-2 minutes, then work
 - **Local only**: No network requests after model download
 - **Model source**: Official HuggingFace repository ([verify here](https://huggingface.co/Xenova/all-MiniLM-L6-v2))
 
-</details>
-
 <details>
 <summary><strong>Performance</strong></summary>
 

package/dist/cli/common.d.ts

@@ -0,0 +1,14 @@
+import { Embedder } from '../embedder/index.js';
+import { VectorStore } from '../vectordb/index.js';
+import type { ResolvedGlobalConfig } from './options.js';
+/**
+ * Create an uninitialized VectorStore from resolved global config.
+ * Callers are responsible for calling initialize() before use.
+ */
+export declare function createVectorStore(config: ResolvedGlobalConfig): VectorStore;
+/**
+ * Create an uninitialized Embedder from resolved global config.
+ * Callers are responsible for managing the Embedder lifecycle.
+ */
+export declare function createEmbedder(config: ResolvedGlobalConfig): Embedder;
+//# sourceMappingURL=common.d.ts.map

package/dist/cli/common.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"common.d.ts","sourceRoot":"","sources":["../../src/cli/common.ts"],"names":[],"mappings":"AAEA,OAAO,EAAE,QAAQ,EAAE,MAAM,sBAAsB,CAAA;AAC/C,OAAO,EAAE,WAAW,EAAE,MAAM,sBAAsB,CAAA;AAClD,OAAO,KAAK,EAAE,oBAAoB,EAAE,MAAM,cAAc,CAAA;AAExD;;;GAGG;AACH,wBAAgB,iBAAiB,CAAC,MAAM,EAAE,oBAAoB,GAAG,WAAW,CAK3E;AAED;;;GAGG;AACH,wBAAgB,cAAc,CAAC,MAAM,EAAE,oBAAoB,GAAG,QAAQ,CAMrE"}

package/dist/cli/common.js

@@ -0,0 +1,25 @@
+// Shared CLI component helpers — factory functions for VectorStore and Embedder
+import { Embedder } from '../embedder/index.js';
+import { VectorStore } from '../vectordb/index.js';
+/**
+ * Create an uninitialized VectorStore from resolved global config.
+ * Callers are responsible for calling initialize() before use.
+ */
+export function createVectorStore(config) {
+    return new VectorStore({
+        dbPath: config.dbPath,
+        tableName: 'chunks',
+    });
+}
+/**
+ * Create an uninitialized Embedder from resolved global config.
+ * Callers are responsible for managing the Embedder lifecycle.
+ */
+export function createEmbedder(config) {
+    return new Embedder({
+        modelPath: config.modelName,
+        batchSize: 16,
+        cacheDir: config.cacheDir,
+    });
+}
+//# sourceMappingURL=common.js.map

package/dist/cli/common.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"common.js","sourceRoot":"","sources":["../../src/cli/common.ts"],"names":[],"mappings":"AAAA,gFAAgF;AAEhF,OAAO,EAAE,QAAQ,EAAE,MAAM,sBAAsB,CAAA;AAC/C,OAAO,EAAE,WAAW,EAAE,MAAM,sBAAsB,CAAA;AAGlD;;;GAGG;AACH,MAAM,UAAU,iBAAiB,CAAC,MAA4B;IAC5D,OAAO,IAAI,WAAW,CAAC;QACrB,MAAM,EAAE,MAAM,CAAC,MAAM;QACrB,SAAS,EAAE,QAAQ;KACpB,CAAC,CAAA;AACJ,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,cAAc,CAAC,MAA4B;IACzD,OAAO,IAAI,QAAQ,CAAC;QAClB,SAAS,EAAE,MAAM,CAAC,SAAS;QAC3B,SAAS,EAAE,EAAE;QACb,QAAQ,EAAE,MAAM,CAAC,QAAQ;KAC1B,CAAC,CAAA;AACJ,CAAC"}

package/dist/cli/delete.d.ts

@@ -0,0 +1,8 @@
+import type { GlobalOptions } from './options.js';
+/**
+ * Run the delete CLI subcommand.
+ * @param args - Arguments after "delete"
+ * @param globalOptions - Global options parsed before the subcommand
+ */
+export declare function runDelete(args: string[], globalOptions?: GlobalOptions): Promise<void>;
+//# sourceMappingURL=delete.d.ts.map

package/dist/cli/delete.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"delete.d.ts","sourceRoot":"","sources":["../../src/cli/delete.ts"],"names":[],"mappings":"AAUA,OAAO,KAAK,EAAE,aAAa,EAAE,MAAM,cAAc,CAAA;AAkFjD;;;;GAIG;AACH,wBAAsB,SAAS,CAAC,IAAI,EAAE,MAAM,EAAE,EAAE,aAAa,GAAE,aAAkB,GAAG,OAAO,CAAC,IAAI,CAAC,CAiGhG"}

package/dist/cli/delete.js

@@ -0,0 +1,164 @@
+// CLI delete subcommand — delete ingested content by file path or source URL
+import { unlink } from 'node:fs/promises';
+import { resolve } from 'node:path';
+import { generateMetaJsonPath, generateRawDataPath, isRawDataPath, } from '../utils/raw-data-utils.js';
+import { createVectorStore } from './common.js';
+import { resolveGlobalConfig, validatePath } from './options.js';
+// ============================================
+// Help
+// ============================================
+const HELP_TEXT = `Usage: mcp-local-rag [global-options] delete [--source <url>] [<file-path>]
+
+Delete ingested content by file path or source URL.
+
+Either <file-path> or --source is required (not both).
+
+Arguments:
+  <file-path>            File path of ingested content to delete
+
+Options:
+  --source <url>         Delete by source URL (for content ingested via ingest_data)
+  -h, --help             Show this help
+
+Global options (must appear before "delete"):
+  --db-path <path>       LanceDB database path
+  --cache-dir <path>     Model cache directory
+  --model-name <name>    Embedding model`;
+/**
+ * Parse delete-specific CLI arguments.
+ * Accepts a positional <file-path>, --source <url>, and -h/--help.
+ * Unknown flags or conflicting args cause exit(1).
+ */
+function parseArgs(args) {
+    let help = false;
+    let source;
+    let filePath;
+    let i = 0;
+    while (i < args.length) {
+        const arg = args[i];
+        if (arg === '-h' || arg === '--help') {
+            help = true;
+            i++;
+        }
+        else if (arg === '--source') {
+            const value = args[++i];
+            if (value === undefined || value.startsWith('-')) {
+                console.error('Missing value for --source');
+                console.error(HELP_TEXT);
+                process.exit(1);
+            }
+            source = value;
+            i++;
+        }
+        else if (arg.startsWith('-')) {
+            console.error(`Unknown option: ${arg}`);
+            console.error(HELP_TEXT);
+            process.exit(1);
+        }
+        else {
+            // Positional argument: file-path
+            filePath = arg;
+            i++;
+        }
+    }
+    const result = { help };
+    if (source !== undefined)
+        result.source = source;
+    if (filePath !== undefined)
+        result.filePath = filePath;
+    return result;
+}
+// ============================================
+// Main Entry Point
+// ============================================
+/**
+ * Run the delete CLI subcommand.
+ * @param args - Arguments after "delete"
+ * @param globalOptions - Global options parsed before the subcommand
+ */
+export async function runDelete(args, globalOptions = {}) {
+    // Parse CLI options
+    const parsed = parseArgs(args);
+    // Handle --help
+    if (parsed.help) {
+        console.error(HELP_TEXT);
+        process.exit(0);
+    }
+    // Validate: either file-path or --source required, not both
+    if (!parsed.filePath && !parsed.source) {
+        console.error('Either <file-path> or --source is required');
+        console.error(HELP_TEXT);
+        process.exit(1);
+    }
+    if (parsed.filePath && parsed.source) {
+        console.error('Cannot specify both <file-path> and --source');
+        console.error(HELP_TEXT);
+        process.exit(1);
+    }
+    // Resolve global config
+    const globalConfig = resolveGlobalConfig(globalOptions);
+    try {
+        // Create and initialize VectorStore (no Embedder needed for delete)
+        const vectorStore = createVectorStore(globalConfig);
+        await vectorStore.initialize();
+        // Determine target file path
+        let targetPath;
+        if (parsed.source) {
+            // Generate raw-data path from source URL
+            targetPath = generateRawDataPath(globalConfig.dbPath, parsed.source, 'markdown');
+        }
+        else {
+            // Use provided file path, resolve to absolute
+            targetPath = resolve(parsed.filePath);
+            // Validate path (reject sensitive system directories)
+            const pathError = validatePath(targetPath, '<file-path>');
+            if (pathError) {
+                console.error(pathError);
+                process.exitCode = 1;
+                return;
+            }
+        }
+        // Delete chunks from VectorStore
+        await vectorStore.deleteChunks(targetPath);
+        // Clean up physical raw-data files if applicable
+        if (isRawDataPath(targetPath)) {
+            try {
+                await unlink(targetPath);
+            }
+            catch (error) {
+                // Ignore ENOENT (file already deleted / never existed)
+                if (!(error instanceof Error) ||
+                    !('code' in error) ||
+                    error.code !== 'ENOENT') {
+                    throw error;
+                }
+            }
+            try {
+                await unlink(generateMetaJsonPath(targetPath));
+            }
+            catch (error) {
+                // Ignore ENOENT
+                if (!(error instanceof Error) ||
+                    !('code' in error) ||
+                    error.code !== 'ENOENT') {
+                    throw error;
+                }
+            }
+        }
+        // Optimize VectorStore after deletion
+        await vectorStore.optimize();
+        // Output result JSON to stdout
+        const result = {
+            filePath: targetPath,
+            deleted: true,
+            timestamp: new Date().toISOString(),
+        };
+        process.stdout.write(JSON.stringify(result));
+    }
+    catch (error) {
+        const reason = error instanceof Error ? error.message : String(error);
+        console.error(`Error: ${reason}`);
+        process.exit(1);
+    }
+}
+//# sourceMappingURL=delete.js.map

package/dist/cli/delete.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"delete.js","sourceRoot":"","sources":["../../src/cli/delete.ts"],"names":[],"mappings":"AAAA,6EAA6E;AAE7E,OAAO,EAAE,MAAM,EAAE,MAAM,kBAAkB,CAAA;AACzC,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAA;AACnC,OAAO,EACL,oBAAoB,EACpB,mBAAmB,EACnB,aAAa,GACd,MAAM,4BAA4B,CAAA;AACnC,OAAO,EAAE,iBAAiB,EAAE,MAAM,aAAa,CAAA;AAE/C,OAAO,EAAE,mBAAmB,EAAE,YAAY,EAAE,MAAM,cAAc,CAAA;AAEhE,+CAA+C;AAC/C,OAAO;AACP,+CAA+C;AAE/C,MAAM,SAAS,GAAG;;;;;;;;;;;;;;;;yCAgBuB,CAAA;AAYzC;;;;GAIG;AACH,SAAS,SAAS,CAAC,IAAc;IAC/B,IAAI,IAAI,GAAG,KAAK,CAAA;IAChB,IAAI,MAA0B,CAAA;IAC9B,IAAI,QAA4B,CAAA;IAEhC,IAAI,CAAC,GAAG,CAAC,CAAA;IACT,OAAO,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;QACvB,MAAM,GAAG,GAAG,IAAI,CAAC,CAAC,CAAE,CAAA;QAEpB,IAAI,GAAG,KAAK,IAAI,IAAI,GAAG,KAAK,QAAQ,EAAE,CAAC;YACrC,IAAI,GAAG,IAAI,CAAA;YACX,CAAC,EAAE,CAAA;QACL,CAAC;aAAM,IAAI,GAAG,KAAK,UAAU,EAAE,CAAC;YAC9B,MAAM,KAAK,GAAG,IAAI,CAAC,EAAE,CAAC,CAAC,CAAA;YACvB,IAAI,KAAK,KAAK,SAAS,IAAI,KAAK,CAAC,UAAU,CAAC,GAAG,CAAC,EAAE,CAAC;gBACjD,OAAO,CAAC,KAAK,CAAC,4BAA4B,CAAC,CAAA;gBAC3C,OAAO,CAAC,KAAK,CAAC,SAAS,CAAC,CAAA;gBACxB,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;YACjB,CAAC;YACD,MAAM,GAAG,KAAK,CAAA;YACd,CAAC,EAAE,CAAA;QACL,CAAC;aAAM,IAAI,GAAG,CAAC,UAAU,CAAC,GAAG,CAAC,EAAE,CAAC;YAC/B,OAAO,CAAC,KAAK,CAAC,mBAAmB,GAAG,EAAE,CAAC,CAAA;YACvC,OAAO,CAAC,KAAK,CAAC,SAAS,CAAC,CAAA;YACxB,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;QACjB,CAAC;aAAM,CAAC;YACN,iCAAiC;YACjC,QAAQ,GAAG,GAAG,CAAA;YACd,CAAC,EAAE,CAAA;QACL,CAAC;IACH,CAAC;IAED,MAAM,MAAM,GAAe,EAAE,IAAI,EAAE,CAAA;IACnC,IAAI,MAAM,KAAK,SAAS;QAAE,MAAM,CAAC,MAAM,GAAG,MAAM,CAAA;IAChD,IAAI,QAAQ,KAAK,SAAS;QAAE,MAAM,CAAC,QAAQ,GAAG,QAAQ,CAAA;IACtD,OAAO,MAAM,CAAA;AACf,CAAC;AAED,+CAA+C;AAC/C,mBAAmB;AACnB,+CAA+C;AAE/C;;;;GAIG;AACH,MAAM,CAAC,KAAK,UAAU,SAAS,CAAC,IAAc,EAAE,gBAA+B,EAAE;IAC/E,oBAAoB;IACpB,MAAM,MAAM,GAAG,SAAS,CAAC,IAAI,CAAC,CAAA;IAE9B,gBAAgB;IAChB,IAAI,MAAM,CAAC,IAAI,EAAE,CAAC;QAChB,OAAO,CAAC,KAAK,CAAC,SAAS,CAAC,CAAA;QACxB,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IACjB,CAAC;IAED,4DAA4D;IAC5D,IAAI,CAAC,MAAM,CAAC,QAAQ,IAAI,CAAC,MAAM,CAAC,MAAM,EAAE,CAAC;QACvC,OAAO,CAAC,KAAK,CAAC,4CAA4C,CAAC,CAAA;QAC3D,OAAO,CAAC,KAAK,CAAC,SAAS,CAAC,CAAA;QACxB,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IACjB,CAAC;IAED,IAAI,MAAM,CAAC,QAAQ,IAAI,MAAM,CAAC,MAAM,EAAE,CAAC;QACrC,OAAO,CAAC,KAAK,CAAC,8CAA8C,CAAC,CAAA;QAC7D,OAAO,CAAC,KAAK,CAAC,SAAS,CAAC,CAAA;QACxB,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IACjB,CAAC;IAED,wBAAwB;IACxB,MAAM,YAAY,GAAG,mBAAmB,CAAC,aAAa,CAAC,CAAA;IAEvD,IAAI,CAAC;QACH,oEAAoE;QACpE,MAAM,WAAW,GAAG,iBAAiB,CAAC,YAAY,CAAC,CAAA;QACnD,MAAM,WAAW,CAAC,UAAU,EAAE,CAAA;QAE9B,6BAA6B;QAC7B,IAAI,UAAkB,CAAA;QAEtB,IAAI,MAAM,CAAC,MAAM,EAAE,CAAC;YAClB,yCAAyC;YACzC,UAAU,GAAG,mBAAmB,CAAC,YAAY,CAAC,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,UAAU,CAAC,CAAA;QAClF,CAAC;aAAM,CAAC;YACN,8CAA8C;YAC9C,UAAU,GAAG,OAAO,CAAC,MAAM,CAAC,QAAS,CAAC,CAAA;YAEtC,sDAAsD;YACtD,MAAM,SAAS,GAAG,YAAY,CAAC,UAAU,EAAE,aAAa,CAAC,CAAA;YACzD,IAAI,SAAS,EAAE,CAAC;gBACd,OAAO,CAAC,KAAK,CAAC,SAAS,CAAC,CAAA;gBACxB,OAAO,CAAC,QAAQ,GAAG,CAAC,CAAA;gBACpB,OAAM;YACR,CAAC;QACH,CAAC;QAED,iCAAiC;QACjC,MAAM,WAAW,CAAC,YAAY,CAAC,UAAU,CAAC,CAAA;QAE1C,iDAAiD;QACjD,IAAI,aAAa,CAAC,UAAU,CAAC,EAAE,CAAC;YAC9B,IAAI,CAAC;gBACH,MAAM,MAAM,CAAC,UAAU,CAAC,CAAA;YAC1B,CAAC;YAAC,OAAO,KAAc,EAAE,CAAC;gBACxB,uDAAuD;gBACvD,IACE,CAAC,CAAC,KAAK,YAAY,KAAK,CAAC;oBACzB,CAAC,CAAC,MAAM,IAAI,KAAK,CAAC;oBACjB,KAA+B,CAAC,IAAI,KAAK,QAAQ,EAClD,CAAC;oBACD,MAAM,KAAK,CAAA;gBACb,CAAC;YACH,CAAC;YAED,IAAI,CAAC;gBACH,MAAM,MAAM,CAAC,oBAAoB,CAAC,UAAU,CAAC,CAAC,CAAA;YAChD,CAAC;YAAC,OAAO,KAAc,EAAE,CAAC;gBACxB,gBAAgB;gBAChB,IACE,CAAC,CAAC,KAAK,YAAY,KAAK,CAAC;oBACzB,CAAC,CAAC,MAAM,IAAI,KAAK,CAAC;oBACjB,KAA+B,CAAC,IAAI,KAAK,QAAQ,EAClD,CAAC;oBACD,MAAM,KAAK,CAAA;gBACb,CAAC;YACH,CAAC;QACH,CAAC;QAED,sCAAsC;QACtC,MAAM,WAAW,CAAC,QAAQ,EAAE,CAAA;QAE5B,+BAA+B;QAC/B,MAAM,MAAM,GAAG;YACb,QAAQ,EAAE,UAAU;YACpB,OAAO,EAAE,IAAI;YACb,SAAS,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;SACpC,CAAA;QACD,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,CAAA;IAC9C,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,MAAM,MAAM,GAAG,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAA;QACrE,OAAO,CAAC,KAAK,CAAC,UAAU,MAAM,EAAE,CAAC,CAAA;QACjC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;IACjB,CAAC;AACH,CAAC"}

package/dist/cli/ingest.d.ts

@@ -1,24 +1,37 @@
-interface CliOptions {
-    dbPath?: string | undefined;
+import type { GlobalOptions, ResolvedGlobalConfig } from './options.js';
+interface IngestConfig {
+    baseDir: string;
+    dbPath: string;
+    cacheDir: string;
+    modelName: string;
+    maxFileSize: number;
+}
+interface IngestCliOptions {
     baseDir?: string | undefined;
-    cacheDir?: string | undefined;
-    modelName?: string | undefined;
     maxFileSize?: number | undefined;
 }
 interface ParsedArgs {
     positional: string | undefined;
-    options: CliOptions;
+    options: IngestCliOptions;
     help: boolean;
 }
 /**
- * Parse CLI arguments into options and a positional path.
- * Flags: --db-path, --base-dir, --cache-dir, --model-name, --max-file-size, -h/--help
+ * Parse ingest-specific CLI arguments into options and a positional path.
+ * Flags: --base-dir, --max-file-size, -h/--help
+ * Unknown flags (including global flags passed after subcommand) cause an error.
  */
 export declare function parseArgs(args: string[]): ParsedArgs;
+/**
+ * Resolve ingest config by merging global config with ingest-specific options.
+ * Ingest-specific: baseDir, maxFileSize (CLI flags > env vars > defaults).
+ * Validates all resolved values before returning.
+ */
+export declare function resolveConfig(globalConfig: ResolvedGlobalConfig, ingestOptions?: IngestCliOptions): IngestConfig;
 /**
  * Run the ingest CLI subcommand.
  * @param args - Arguments after "ingest" (e.g., option flags and file/directory path)
+ * @param globalOptions - Global options parsed before the subcommand
  */
-export declare function runIngest(args: string[]): Promise<void>;
+export declare function runIngest(args: string[], globalOptions?: GlobalOptions): Promise<void>;
 export {};
 //# sourceMappingURL=ingest.d.ts.map

package/dist/cli/ingest.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"ingest.d.ts","sourceRoot":"","sources":["../../src/cli/ingest.ts"],"names":[],"mappings":"AA8BA,UAAU,UAAU;IAClB,MAAM,CAAC,EAAE,MAAM,GAAG,SAAS,CAAA;IAC3B,OAAO,CAAC,EAAE,MAAM,GAAG,SAAS,CAAA;IAC5B,QAAQ,CAAC,EAAE,MAAM,GAAG,SAAS,CAAA;IAC7B,SAAS,CAAC,EAAE,MAAM,GAAG,SAAS,CAAA;IAC9B,WAAW,CAAC,EAAE,MAAM,GAAG,SAAS,CAAA;CACjC;AAED,UAAU,UAAU;IAClB,UAAU,EAAE,MAAM,GAAG,SAAS,CAAA;IAC9B,OAAO,EAAE,UAAU,CAAA;IACnB,IAAI,EAAE,OAAO,CAAA;CACd;AAiCD;;;GAGG;AACH,wBAAgB,SAAS,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,UAAU,CAiDpD;AAmJD;;;GAGG;AACH,wBAAsB,SAAS,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC,CAgG7D"}
+{"version":3,"file":"ingest.d.ts","sourceRoot":"","sources":["../../src/cli/ingest.ts"],"names":[],"mappings":"AAWA,OAAO,KAAK,EAAE,aAAa,EAAE,oBAAoB,EAAE,MAAM,cAAc,CAAA;AAavE,UAAU,YAAY;IACpB,OAAO,EAAE,MAAM,CAAA;IACf,MAAM,EAAE,MAAM,CAAA;IACd,QAAQ,EAAE,MAAM,CAAA;IAChB,SAAS,EAAE,MAAM,CAAA;IACjB,WAAW,EAAE,MAAM,CAAA;CACpB;AAQD,UAAU,gBAAgB;IACxB,OAAO,CAAC,EAAE,MAAM,GAAG,SAAS,CAAA;IAC5B,WAAW,CAAC,EAAE,MAAM,GAAG,SAAS,CAAA;CACjC;AAED,UAAU,UAAU;IAClB,UAAU,EAAE,MAAM,GAAG,SAAS,CAAA;IAC9B,OAAO,EAAE,gBAAgB,CAAA;IACzB,IAAI,EAAE,OAAO,CAAA;CACd;AAgCD;;;;GAIG;AACH,wBAAgB,SAAS,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,UAAU,CAwDpD;AAMD;;;;GAIG;AACH,wBAAgB,aAAa,CAC3B,YAAY,EAAE,oBAAoB,EAClC,aAAa,GAAE,gBAAqB,GACnC,YAAY,CA6Bd;AAuJD;;;;GAIG;AACH,wBAAsB,SAAS,CAAC,IAAI,EAAE,MAAM,EAAE,EAAE,aAAa,GAAE,aAAkB,GAAG,OAAO,CAAC,IAAI,CAAC,CA0FhG"}