Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server
dev.agentdiet/rag-dietagentdiet.devstreamable-httpv7.0.17Scanned
Strips 85-97% of HTML/DOM boilerplate into clean Markdown, skeletons, or BM25-ranked RAG chunks.
4 tools: 2 change things and 2 only read.
- Reads private data · 0
- Reads text others write · 3
- Can send data out · 3
- No findings
Tools listed without sign-in. Remote posture
With Anlyon
2 tools here change things. Govern them.
- Anlyon lists the tools and adopts none.
- Each tool you adopt gets approvals and receipts.
- Your agent never holds the key.
A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.
What the Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server can do
The Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server is published by agentdiet.dev, a domain the MCP Registry verified. It has 4 tools. It can create or change data (2 tools: recon_webpage_skeleton and extract_structured_data), but cannot delete data. 2 tools only read.
3 tools bring in text other people can write (distill_webpage, recon_webpage_skeleton and extract_structured_data); 3 tools can send data out of your system (distill_webpage, recon_webpage_skeleton and extract_structured_data).
None of the fixed checks matched anything in its tool definitions. That is not a safety guarantee.
Its remote endpoint lists its tools without sign-in.
Last scanned Nov 17, 2026, version 7.0.17.
Tools
Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.
4 tools
Changes things 2Claude Code: ask
extract_structured_dataExtract machine-readable structured data from a web page: Schema.org JSON-LD objects, OpenGraph metadata, HTML tables parsed as JSON row arrays, fenced code blocks, section heading…Reads text others writeCan send data outNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Description, as the server wrote it Extract machine-readable structured data from a web page: Schema.org JSON-LD objects, OpenGraph metadata, HTML tables parsed as JSON row arrays, fenced code blocks, section headings, and links. WHEN TO USE: Call this tool for pricing tables, API specifications, product metadata, or code snippets; use distill_webpage for narrative Markdown or recon_webpage_skeleton for a lightweight heading outline. WHEN NOT TO USE: Do not use for narrative prose extraction or binary files. PARAMETER INTERACTIONS: Provide either url or html (when html is supplied, url is not fetched and only labels the cache key). Omit agent_id and tx_hash during your 5 free calls. Read-only and idempotent; results are cached for 1 hour (<5ms cache hit). $0.001 per call ($0.0005 cached). Pricing: the first 5 calls per agent_id are free and failed calls are never charged. After 5 calls, the tool returns a payment-required quote (not a fatal error) with a $2 Stripe top-up link and Base/Solana USDC addresses; complete payment, then retry with your agd_pass_ key in agent_id or pass tx_hash. Example: extract_structured_data(url='https://docs.stripe.com/api')
Annotations the server declares
- title: Extract structured data
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true
recon_webpage_skeletonExtract an ultra-compact structural outline (under ~200 tokens) of a web page listing its H1-H4 headings and exact token count per section without returning body prose.Reads text others writeCan send data outNo findings
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Description, as the server wrote it Extract an ultra-compact structural outline (under ~200 tokens) of a web page listing its H1-H4 headings and exact token count per section without returning body prose. WHEN TO USE: Call this tool first on long or unfamiliar pages to inspect structure and token cost before calling distill_webpage (with mode='budget' and a focused query). WHEN NOT TO USE: Do not use when you already need full Markdown prose (use distill_webpage) or structured tables and JSON-LD (use extract_structured_data). PARAMETER INTERACTIONS: Provide either url or html (when html is supplied, url is not fetched and only labels the cache key). Omit agent_id and tx_hash during your 5 free calls. Read-only and idempotent; results are cached for 1 hour (<5ms cache hit). $0.0002 per call. Pricing: the first 5 calls per agent_id are free and failed calls are never charged. After 5 calls, the tool returns a payment-required quote (not a fatal error) with a $2 Stripe top-up link and Base/Solana USDC addresses; complete payment, then retry with your agd_pass_ key in agent_id or pass tx_hash. Example: recon_webpage_skeleton(url='https://docs.stripe.com/api')
Annotations the server declares
- title: Recon webpage skeleton
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true
Only reads 2Show 2 read-only tools
distill_webpageFetch a public HTTP(S) web page (or accept raw HTML) and distill its main prose into clean Markdown, stripping 85-97% of HTML/JS/CSS boilerplate.Reads text others writeCan send data outNo findings
Why it is in this group
- The description's first verb is "fetch": a word for reading.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
- untrusted-inputThe name "distill_webpage" says "webpage": it reads web pages, which anyone can write.
- untrusted-inputThe description's first sentence says "fetch" and "http": it reads web pages, which anyone can write.
- untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Claude Code: ask, because it is tagged outbound.
Description, as the server wrote it Fetch a public HTTP(S) web page (or accept raw HTML) and distill its main prose into clean Markdown, stripping 85-97% of HTML/JS/CSS boilerplate. WHEN TO USE: Call this tool to read full or query-filtered article and documentation text; call recon_webpage_skeleton first to preview headings and token counts on unfamiliar pages, or extract_structured_data when you need JSON-LD, HTML tables, or code blocks. WHEN NOT TO USE: Do not use for binary PDFs, images, or login-walled pages. PARAMETER INTERACTIONS: Provide either url or html (when html is supplied, url is not fetched and only labels the cache key). When mode='markdown' (default), query and max_tokens are ignored; set mode='budget' to rank sections by query (BM25) and cap output at max_tokens. Omit agent_id and tx_hash during your 5 free calls. Read-only and idempotent; results are cached for 1 hour (<5ms cache hit). $0.001 per call ($0.0005 cached). Pricing: the first 5 calls per agent_id are free and failed calls are never charged. After 5 calls, the tool returns a payment-required quote (not a fatal error) with a $2 Stripe top-up link and Base/Solana USDC addresses; complete payment, then retry with your agd_pass_ key in agent_id or pass tx_hash. Example: distill_webpage(url='https://docs.stripe.com/api')
Annotations the server declares
- title: Distill webpage
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true
get_agent_wallet_statusInspect an agent's account ledger: remaining free trial calls (out of 5), prepaid USD balance, cumulative spend, total API calls, and total LLM tokens saved.No findings
Why it is in this group
- The name "get_agent_wallet_status" says "get": a word for reading.
Description, as the server wrote it Inspect an agent's account ledger: remaining free trial calls (out of 5), prepaid USD balance, cumulative spend, total API calls, and total LLM tokens saved. WHEN TO USE: Call before a batch of paid tool executions or after completing a Stripe or USDC top-up to confirm your balance. WHEN NOT TO USE: Do not use to process web content, execute marketplace tools, or submit payments; use the other tools on this server for execution. PARAMETER INTERACTIONS: Omit agent_id to inspect the default trial account derived from your client IP, or pass your explicit agent_id or agd_pass_ session key to check a specific funded wallet. Free and read-only: never charges a fee or consumes a free trial call. Example: get_agent_wallet_status(agent_id='example')
Annotations the server declares
- title: Get agent wallet status
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.
Exfil surface
Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.
No private-data, untrusted-input, outbound path in this server alone: no tool is tagged private-data.
- Reads private data0
- Reads files, mail, messages, databases, secrets, calendars or repositories.
- Reads text others write3
- Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
- Can send data out3
- Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.
Remote posture
What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.
Looked at Nov 20, 2026, 19:29 UTC, unchanged since Oct 22, 2026. No credentials were sent and no tool was called. What is looked at.
- Reachable
- yes
- HTTP status of an anonymous initialize
- 200
- tools/list served without credentials
- yes
- Tools that change things, listed without credentials
- 2
- Protocol version
- 2025-06-18
- WWW-Authenticate on 401
- not observed
- resource_metadata in WWW-Authenticate
- not observed
- RFC 9728 protected resource metadata
- https://rag-diet.agentdiet.dev/.well-known/oauth-protected-resource/mcp
- Authorization servers
- https://rag-diet.agentdiet.dev
- Legacy HTTP+SSE transport
- no
- Dynamic client registration
- yes
- Client ID metadata documents
- yes
- WarnThe whole server
The endpoint served tools/list without credentials, and the list has 2 tools that change things (2 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://rag-diet.agentdiet.dev/mcp: tools/list returned 4 tools to an anonymous client, including recon_webpage_skeleton (write), extract_structured_data (write)
Findings
Matches of fixed checks, with the text they matched.
No check matched anything in these tool definitions. That is not a safety guarantee.
History
Changes to its tool definitions and to its remote endpoint, newest first.
No change seen yet. Its tools are rescanned every day, and a remote endpoint is looked at once a week.
Install
Pinned to version 7.0.17. Values in angle brackets are yours to fill in.
claude mcp add --transport http rag-diet https://rag-diet.agentdiet.dev/mcpclaude mcp add --transport http rag-diet https://rag-diet.agentdiet.dev/mcpRun in a terminal.
Claude Code permissions
Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.
{
"permissions": {
"allow": [
"mcp__rag-diet__get_agent_wallet_status"
],
"ask": [
"mcp__rag-diet__distill_webpage",
"mcp__rag-diet__recon_webpage_skeleton",
"mcp__rag-diet__extract_structured_data"
],
"deny": []
}
}Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as rag-diet. Review it before you use it.
Capability tags behind 3 rules
- mcp__rag-diet__distill_webpage untrusted-input, outbound
- mcp__rag-diet__recon_webpage_skeleton untrusted-input, outbound
- mcp__rag-diet__extract_structured_data untrusted-input, outbound
Pin
Keep the decision: fail CI when the tools change, follow every change, or show a badge.
Pin in CI
Fail a build when the tools you reviewed are not the tools the server serves today.
curl -s "https://radar.anlyon.com/api/v1/servers/dev.agentdiet/rag-diet/verify?hash=792a51205cb35efb2c988995930ec23bd0d2aa12d8ab5bda3a0af4c590c10a68" | jq -e '.changed == false'Pinned hash 792a51205cb3, from Nov 17, 2026, 22:17 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/dev.agentdiet/rag-diet/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/dev.agentdiet/rag-diet)Source: MCP Registry · repository. Synced Oct 21, 2026, 18:11 UTC.
Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server safe?
- No scan can promise that. What its definitions show: 4 tools, 2 of which can change, send or delete something, and 0 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.
Can the Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server delete data?
- No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.
Can the Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server do things that cannot be undone?
- No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.
What tools does the Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server have?
- 4 tools: 2 write, 2 read. recon_webpage_skeleton, extract_structured_data, distill_webpage and get_agent_wallet_status.
Does the Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server send data to outside services?
- 3 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): distill_webpage, recon_webpage_skeleton and extract_structured_data.
Has the Imperium RAG-Diet — Web & Document Context Distiller (85-97% Token Savings) MCP server changed its tools?
- Not since Anlyon Radar started tracking it. Last scanned Nov 17, 2026.