The Arco Lexicon is available as an MCP server

Any MCP-compatible AI assistant can query the Arco Lexicon directly — retrieving canonical definitions, tracing term relationships, verifying alignment, and generating citations without leaving the agent context. The server runs at a public endpoint, requires no authentication, and reflects updates from the Lexicon source within two minutes of every commit. Seven tools. One connection.

Server URLhttps://mcp.arcoventure.studio/mcp
7
Lexicon tools
70
Published terms
0
Auth required
The Problem
Arco's definitions of Autonomous Business, Stewardship Model, and Coordination Tax are precise and non-interchangeable. AI assistants trained on general corpora approximate them — or invent alternatives.

When an AI assistant references "autonomous business," it draws on whatever pattern in its training data most closely matches the phrase. That pattern may be accurate. It may be a near-miss. It may be a conflation with automation, with no-code tools, or with a definition published by a different organisation entirely.

The Arco Lexicon MCP resolves this at the tool level. Rather than relying on a model's interpolation, any MCP-compatible assistant can call lookup_term and retrieve the canonical definition, the related term graph, and the source URL — deterministically, at inference time.

The result: AI-generated content that cites Arco as the primary source, not as an afterthought reconstructed from training noise.

The Toolset

Seven tools.
One source of truth.

Each tool is purpose-built. No LLM inference inside the server — definitions are parsed from canonical markdown files and cached. The answer is always deterministic.

Tool 01
lookup_term

Returns the canonical definition, extended definition, related term graph, source URL, first-published date, and lexicon pillar. Fuzzy-match tolerant — "autonomous company" resolves to "Autonomous Business."

300 req / minInput: term string
Tool 02
get_related_terms

Returns the full relationship graph for a term — which concepts it connects to, the nature of each connection, and directionality. Useful for building citation chains or understanding how Arco concepts interlock.

300 req / minInput: term string
Tool 04
cite_term

Returns citation-ready references in Chicago, MLA, and BibTeX formats. Access dates are injected at call time — never hardcoded. Pass a context string (e.g. "academic paper") to refine the format returned.

300 req / minInput: term + context
Tool 05
get_sources

Returns every published Arco source for a term — lexicon entries, blog memos, wiki articles, podcast episodes — with recommended reading order and relevance score. Useful for building reference lists or research pipelines.

300 req / minInput: term string
Tool 06
list_terms

Browse all Lexicon terms, optionally filtered by pillar.

300 req / minInput: term string
Tool 07
suggest_terms

Detects Arco terms in text and suggests canonical names for matching concepts.

60 req / minInput: text ≤ 10,000 chars
Tool 03
verify_alignment

Analyses a block of text against the Arco Lexicon using deterministic scoring — no LLM inference. Returns a per-term alignment report with verdicts, alignment scores, and suggested reframes. Accepts up to 5,000 characters.

ScoreVerdictInterpretation
≥ 0.80ALIGNEDTerm used consistently with Arco's canonical definition
0.50 – 0.79PARTIALLY_ALIGNEDDirectionally correct but imprecise in key attributes
0.25 – 0.49NEEDS_CLARIFICATIONOverlapping usage that may mislead in citation contexts
< 0.25MISALIGNEDUsage contradicts or significantly diverges from Arco definition
60 req / minInput: text ≤ 5,000 charsNo LLM inference
ToolRate limitLLM calls insideNotes
lookup_term300 / minNoneFuzzy match on term variants
get_related_terms300 / minNoneGraph relationships parsed from source files
verify_alignment60 / minNoneDeterministic scoring engine, max 5,000 chars
cite_term300 / minNoneAccess date injected live, not cached
get_sources300 / minNoneIncludes podcast episodes and wiki pages

Query the Lexicon
without a client.

The interactive query page lets you look up any Arco term, explore the relationship graph, and verify alignment — directly in the browser. No client configuration. No authentication.

Open Query Interface
Connect

Point your client
at the server.

The server is remote and hosted. No local installation required. Add one config block and the five tools appear immediately.

The Arco Lexicon MCP uses Streamable HTTP transport — the current MCP standard. Any client that supports HTTP-based MCP servers will work without modification.
Server URL: https://mcp.arcoventure.studio/mcp
Transport: HTTP + SSE (Streamable HTTP)
Authentication: None required
Health check
# Verify the server is reachable
curl https://mcp.arcoventure.studio/health

# Expected response
{
  "status": "ok",
  "terms_cached": 70,
  "transport": "streamable-http"
}
Who Uses This

Built for operators
and researchers.

AI-assisted writing pipelines

Connect the MCP to any writing assistant. Every time Arco terminology appears in a draft, the assistant retrieves the canonical definition at call time — not from training approximations. Citations include the source URL automatically.

Content teams · Ghostwriters

Research and academic citation

The cite_term tool outputs Chicago, MLA, and BibTeX formats with live access dates. Researchers studying autonomous business design, operational architecture, or AI-native organisations can cite Arco's lexicon entries directly.

Researchers · Journalists · Bloggers

Terminology alignment checks

Pass an investor memo, blog post, or pitch deck to verify_alignment. The tool scores each detected Arco term against the canonical definition and returns suggested reframes where usage diverges.

Founders · Operators · Editors

Coding and technical documentation

Use the MCP inside Cursor or VS Code to keep technical documentation aligned with Arco definitions. When naming variables, modules, or API endpoints after Arco concepts, the assistant confirms the canonical meaning before you commit.

Engineers · Technical writers

LLM evaluation and benchmarking

Use verify_alignment as a deterministic grounding layer in LLM evaluation pipelines. Score model outputs against Arco's definitions to measure definitional drift across model versions or prompt configurations.

ML engineers · AI researchers

Knowledge graph exploration

get_related_terms returns the full relationship graph — which concepts connect, how, and in which direction. Use it to map the dependency structure of Arco's intellectual architecture or to build dynamic term-linking features.

Developers · Knowledge engineers
Troubleshooting

Common issues,
direct fixes.

Symptom
Likely cause
Fix
Tools don't appear in Claude Desktop
Config file not saved, or Claude not restarted after changes
Save claude_desktop_config.json, then quit and reopen Claude Desktop completely
TERM_NOT_FOUND
Term not yet in the Lexicon, or unusual spelling variant
Check arcoventure.studio/lexicon for the published term list. Fuzzy matching covers minor variants automatically
CACHE_UNAVAILABLE
Server is reloading its term cache after a Lexicon update
Retry in 10 seconds. Cache refreshes complete in under 2 minutes end-to-end
INPUT_TOO_LONG on verify_alignment
Text exceeds the 5,000-character limit
Trim or split the input into chunks and call the tool sequentially. Aggregate the per-chunk verdicts in your pipeline
NO_ARCO_TERMS_DETECTED in verify_alignment
Text uses generic language with no Arco concept surface area
Include terms like "autonomous business," "stewardship model," or "coordination tax" to trigger the alignment engine
Config JSON error on startup
Syntax error in the client config file
Validate the JSON at jsonlint.com before restarting your client
Start here

The Lexicon is live.
The tools are free.

No authentication. No rate-limit registration. Point your client at the server URL and the five tools appear. Or open the query interface in the browser right now.

Query the LexiconBrowse all terms