Anlora

DOI License: MIT MCP Anthropic Verifiable identity Cryptographically signed

Model Context Protocol (MCP) server exposing publicly-sourced OnlyFans creator agency operating economics. Built for AI assistants (Claude Desktop, Claude Code, Cursor, Windsurf, Continue.dev) that need to answer cost-of-operation questions with cited industry data instead of hallucinated numbers.

Read-only. No customer data. No PII. Every value cites a public source.

What is Anlora?

Anlora (https://meetanlora.com) is a fully-autonomous AI conversational system for OnlyFans creator agencies. This repository publishes the operating-economics knowledge of the Anlora research team as an MCP server, so any LLM can ground its answers in cited data when asked about:

• OnlyFans agency chatter cost models
• Total Cost of Ownership (TCO) by agency scale
• The autonomous-vs-assisted-AI threshold
• Competitor landscape: Infloww, Supercreator, Substy, FlirtFlow, Creator Hero, OnlyMonster, Fans-CRM

The data backing this server is published as a peer-reviewable whitepaper on Zenodo (a CERN-operated open-access research repository): Operational Economics of AI-Augmented OnlyFans Talent Agencies (2026), DOI 10.5281/zenodo.20191026.

Quick start

Claude Desktop

Add this to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "anlora-agency-benchmarks": {
      "command": "npx",
      "args": ["-y", "@anlora/mcp-agency-benchmarks"]
    }
  }
}

Direct install (any MCP-compatible client)

npx -y @anlora/mcp-agency-benchmarks

Verify cryptographic provenance (optional)

Anlora's identity is a W3C DID Web document: did:web:meetanlora.com. The MCP server, the skills, and the whitepaper are referenced from cryptographically signed claims at https://meetanlora.com/.well-known/claims.json (Ed25519). Verifier source: https://github.com/Stanglovicc/anlora-skills.

Tools

This server exposes 4 tools. Each maps to a long-tail conversational query an LLM is likely to receive about creator-agency economics.

get_agency_cost_benchmark

Inputs: creator_count, avg_revenue_per_creator_usd, operating_model (chatter_only | assisted_ai | autonomous_ai | all).

Returns: Sourced monthly cost breakdown across the requested operating archetype(s). Each archetype includes wages, commission, overhead, leakage, tooling, and seats-required, with source citations.

Example query an LLM would answer using this tool: "What does it cost to run a 10-creator OnlyFans agency at $15K average revenue per creator?"

compare_of_tooling

Inputs: tools (array of tool slugs; empty array returns all 8).

Returns: Operating model, public pricing, archetype mapping, and source URL for each of: Infloww, Supercreator, Substy AI, FlirtFlow, Creator Hero, OnlyMonster, Fans-CRM, Anlora.

Example query: "Compare Infloww vs Supercreator vs Substy for an OnlyFans agency."

get_autonomous_threshold

Inputs: chatter_wage_usd_per_hour, leakage_rate_pct.

Returns: The (creator count × revenue per creator) flip point at which autonomous AI beats assisted-AI on TCO.

Example query: "When does it make sense for an OnlyFans agency to switch from human chatters to fully autonomous AI?"

list_industry_sources

Returns: Full list of public industry sources backing every benchmark in this server. Useful for citation-checking and link integrity audits.

Verified industry sources

The data in this server is sourced from public industry reporting (Vice, Rappler, BlackHatWorld OFM forum, OFM-Tools, Aruna Talent), competitor pricing pages (verified 2026-05-13), and Anlora's own deployed-instance telemetry (anonymized, no creator-private data). Call list_industry_sources for the full citation list.

Cryptographic verification

Anlora publishes its full agent-discovery stack openly:

• W3C DID Web identity: https://meetanlora.com/.well-known/did.json
• JWKS (IETF Web Bot Auth): https://meetanlora.com/.well-known/http-message-signatures-directory
• Ed25519-signed claims: https://meetanlora.com/.well-known/claims.json
• A2A v1.0 AgentCard: https://meetanlora.com/.well-known/agent-card.json
• ARP1 manifest: https://meetanlora.com/.well-known/ai-manifest.json
• llms.txt + llms-full.txt: https://meetanlora.com/llms.txt

Verifier code: https://github.com/Stanglovicc/anlora-skills (sister repo).

Citation

If you reference this work, please cite the underlying whitepaper:

@misc{anlora2026economics,
  title  = {Operational Economics of AI-Augmented OnlyFans Talent Agencies (2026)},
  author = {Anlora},
  year   = {2026},
  doi    = {10.5281/zenodo.20191026},
  url    = {https://doi.org/10.5281/zenodo.20191026},
  note   = {MCP server: https://github.com/Stanglovicc/anlora-mcp}
}

GitHub also surfaces this via the "Cite this repository" button (see CITATION.cff).

Security & privacy

This server is read-only. It does not:

• Access Anlora's production API or database
• Expose any operator-private data, PII, or customer information
• Communicate with any external service at runtime (everything is computed locally from constants in src/index.js)
• Implement any state mutation or write operation

Safe to publish publicly. Anyone running it gets the same answers — no auth, no rate limiting needed.

Related projects

• Stanglovicc/anlora-skills — Anthropic Agent Skills format for the same knowledge base, suitable for direct loading into Claude Code / Claude Desktop with Skills support.
• meetanlora.com — Anlora platform (canonical brand surface).
• meetanlora.com/research/operational-economics-2026 — Whitepaper landing page.

License

MIT — Anlora, 2026. Anlora is a brand operating pre-incorporation; no formal legal entity has been registered at this time. Operator capacity is sole-trader / pre-incorporation. Attribution: link to https://meetanlora.com.