Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/groupdocs-metadata/groupdocs.metadata.mcp/agents-mdgit clone --depth 1 https://github.com/groupdocs-metadata/GroupDocs.Metadata.McpWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01689 | $0.01689 |
| Opus 5 | $0.00844 | $0.00844 |
| Sonnet 5 | $0.00338 | $0.00338 |
| Haiku 4.5 | $0.00169 | $0.00169 |
Grade A, and why
GroupDocs.Metadata.Mcp AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Guide for AI coding agents
Brief orientation for AI coding agents (Claude Code, Copilot, Cursor, Aider, Amp, Codex) working in this repository.
What this repo is
A standalone MCP server for GroupDocs.Metadata for .NET — exposes document metadata operations as AI-callable tools via the Model Context Protocol.
Published to NuGet as GroupDocs.Metadata.Mcp with the McpServer package type, and to ghcr.io/groupdocs-metadata/metadata-net-mcp + docker.io/groupdocs/metadata-net-mcp as a container image.
MCP tools exposed
| Tool | Description |
|---|---|
ReadMetadata |
Extract all metadata (author, title, dates, EXIF, XMP, IPTC, custom properties) from a document and return grouped JSON |
SearchMetadata |
Return only the properties matching a category / nameContains / valueContains filter (thin predicate layer over FindProperties) |
WriteMetadata |
Set/add a property (Author, Title, Subject, Keywords, Comments, Copyright, Company, Manager) via the tag taxonomy and save the file |
RemoveMetadata |
Strip all metadata (Sanitize) by default, or only specific categories (RemoveProperties by tag) |
GetDocumentInfo |
Return file type, MIME type, page count, byte size, and encryption status as JSON (lightweight structural check; no metadata dump) |
All tools accept FileInput (resolved via IFileResolver) and an optional password for protected documents.
Error contract. Each tool wraps its engine call in try/catch and, on failure, returns a descriptive string instead of MCP's opaque "An error occurred invoking '<tool>'". The text starts with a per-tool prefix — Metadata read failed for '<file>': …, Metadata search failed for …, Metadata write failed for …, Metadata removal failed for …, Document-info lookup failed for … — followed by the exception type, message, and inner-exception chain. The shared formatter lives in Tools/ToolError.cs. In evaluation mode RemoveMetadata surfaces the engine's "Could not save the file. Evaluation only." through this same path (so IsError stays false). File-not-found / resolution errors are raised before the try and propagate as normal MCP errors. When adding a tool, keep this contract.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 109 lines · 1,689 tokens per session scan A 827cc9f97d35
GroupDocs.Metadata.Mcp AGENTS.md is an instructions file published in the GitHub repository groupdocs-metadata/GroupDocs.Metadata.Mcp (1 stars, last pushed 1mo ago), licensed MIT. It adds 1,689 tokens to every session, about $0.0084 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
GroupDocs.Signature.Mcp AGENTS.md
AGENTS.md instructions for groupdocs-signature/GroupDocs.Signature.Mcp, covering agents.md — guide for ai coding agents, what this repo is, mcp tools exposed, folder layout and dependencies.
GroupDocs.Parser.Mcp AGENTS.md
AGENTS.md instructions for groupdocs-parser/GroupDocs.Parser.Mcp, covering agents.md — guide for ai coding agents, what this repo is, mcp tools exposed, folder layout and dependencies.
GroupDocs.Redaction.Mcp AGENTS.md
AGENTS.md instructions for groupdocs-redaction/GroupDocs.Redaction.Mcp, covering agents.md — guide for ai coding agents, what this repo is, mcp tools exposed, folder layout and dependencies.
GroupDocs.Watermark.Mcp AGENTS.md
AGENTS.md instructions for groupdocs-watermark/GroupDocs.Watermark.Mcp, covering agents.md — guide for ai coding agents, what this repo is, mcp tools exposed, folder layout and dependencies.
GroupDocs.Viewer.Mcp AGENTS.md
AGENTS.md instructions for groupdocs-viewer/GroupDocs.Viewer.Mcp, covering agents.md — guide for ai coding agents, what this repo is, mcp tools exposed, folder layout and dependencies.
GroupDocs.Comparison.Mcp AGENTS.md
AGENTS.md instructions for groupdocs-comparison/GroupDocs.Comparison.Mcp, covering agents.md — guide for ai coding agents, what this repo is, mcp tools exposed, folder layout and dependencies.