# Troubleshooting & FAQ Solutions to the most common GroupDocs.Parser MCP server issues — server not appearing in the client, startup failures, missing native dependencies, and first-launch timeouts. Solutions to the most common GroupDocs.Parser MCP server issues — server not appearing in the client, startup failures, missing native dependencies, and first-launch timeouts. {{< alert style="info" >}} **Platform-specific troubleshooting:** this product runs as a Docker container, so runtime problems are Docker problems — the daemon, volume mounts, and the first image pull. See [Troubleshooting (.NET)]({{< ref "parser/net/mcp/troubleshooting.md" >}}). The issues on this page apply to every platform. {{< /alert >}} ## Why is my MCP server not showing up in Claude Desktop? 1. **Restart the client** — every client reads its MCP config only at startup. 2. Check the config file location for your OS ([per-client reference]({{< ref "parser/net/mcp/install-in-ai-clients.md" >}})) and that the entry sits under the right root key (`mcpServers` for Claude Desktop/Cursor/Windsurf, `servers` for VS Code/VS 2022). 3. Validate the JSON — a trailing comma silently breaks the whole file. If you used the [installer](https://github.com/groupdocs/GroupDocs.Mcp.Installer), a timestamped `.bak` of your previous config sits next to the file for comparison. ## The first tool call is slow or fails once, then works A **cold cache**: on the very first use the server's package or image is still downloading while the client is already waiting on the connection. Warming it once fixes it for good — the exact command depends on your build: [.NET]({{< ref "parser/net/mcp/troubleshooting.md" >}}#the-first-tool-call-is-slow-or-times-out-once-and-then-works). ## The server fails to start, or a runtime dependency is missing These are properties of the build you run rather than of MCP, so the fixes live with the platform: | Symptom | Where the fix is | |---|---| | "docker daemon not reachable" | [.NET troubleshooting]({{< ref "parser/net/mcp/troubleshooting.md" >}}#docker-daemon-not-reachable) — start Docker Desktop or `dockerd` | | The storage folder looks empty to the agent | [.NET troubleshooting]({{< ref "parser/net/mcp/troubleshooting.md" >}}#the-agent-says-the-storage-folder-is-empty) — the volume did not mount what you expected | | First call times out, later calls work | [.NET troubleshooting]({{< ref "parser/net/mcp/troubleshooting.md" >}}#the-first-tool-call-is-slow-or-times-out-once-and-then-works) — the image is large; pull it once | | You were told to run `dnx` | There is no NuGet package for this product — [why]({{< ref "parser/net/mcp/troubleshooting.md" >}}#why-is-there-no-dnx-command-for-this-product) | ## The agent says a file does not exist Pass the **file name**, not a full path from your machine: the server resolves names inside its configured storage folder. When a name is not found the tool responds with the list of files it can see, so the agent can correct itself — check that list against [`GROUPDOCS_MCP_STORAGE_PATH`]({{< ref "parser/net/mcp/configuration.md" >}}). ## Why is there no `dnx` command for Parser? Because there is no NuGet package. GroupDocs.Parser's engine embeds around **234 MB of ONNX models**, which puts the packed tool over nuget.org's 250 MB limit — so this product ships **via Docker only**. Every install page here uses `docker run`; when a slimmer engine ships, the NuGet channel returns. See [Install for .NET]({{< ref "parser/net/mcp/_index.md" >}}). ## Extraction or conversion — which do I need? **Extraction** pulls values *out* of a document: text, tables, barcodes, metadata. **Conversion** turns a document into a different format, preserving layout. If you want a DOCX as a PDF, that is the [GroupDocs.Conversion MCP server](https://docs.groupdocs.com/conversion/mcp.md); if you want the numbers out of the table on page 3, it is this one. Many pipelines use both. ## Does it OCR scanned documents? Text extraction works on the text layer. The engine's models do power **barcode and QR detection on rasterized content**, so [`extract_barcodes`]({{< ref "parser/mcp/tools-reference/extract-barcodes.md" >}}) works on scans where [`extract_text`]({{< ref "parser/mcp/tools-reference/extract-text.md" >}}) returns nothing. An empty text extraction on a scanned page means "no text layer", not "empty page". ## Why was my text truncated? [`extract_text`]({{< ref "parser/mcp/tools-reference/extract-text.md" >}}) truncates very large outputs and marks where it stopped, so a 900-page book does not arrive as one response. Extract **per page** (`page: 12`) when you need completeness, and loop. ## Markdown or JSON for tables? `format: "markdown"` (the default) renders as a real table in chat, VS Code, and GitHub — best for looking at data. `format: "json"` returns rows as structured data — best when the agent will compute with it or write it somewhere. Ask for the one that matches what happens next. ## Verifying an installation end-to-end Ask your agent *"list your GroupDocs parser tools and the license status"* — it should name `extract_text`, `extract_tables`, `extract_metadata`, `extract_images`, `extract_barcodes`, `get_document_info`, `get_license_status`. For a scripted check that performs the real MCP handshake and a live call through the engine, see [verifying a .NET installation]({{< ref "parser/net/mcp/troubleshooting.md" >}}#verifying-an-installation-end-to-end). ## Still stuck? Post your config (redact license paths) and the client name in the [Parser forum](https://forum.groupdocs.com/c/parser/17). Bugs: [GitHub issues](https://github.com/groupdocs-parser/GroupDocs.Parser.Mcp/issues).