Skip to content
Back to search

mecenai

AI sales concierge for artists and galleries — answers collectors 24/7 in their language, books viewings, hands off to a human.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/monovarton--mecenai/badge.svg)](/listings/monovarton--mecenai)

HTML

<a href="/listings/monovarton--mecenai"><img src="/api/listings/monovarton--mecenai/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 108 lines
  1. # Mecenai
  2. **An AI sales concierge for artists and galleries.**
  3. Mecenai answers collectors 24/7 in their own language, knows every artwork in the catalogue, books viewings and hands the conversation to a human at the right moment.
  4. > *Mecenai* — from *mecenate*, a patron of the arts, + AI. It connects artists with the collectors who will support them.
  5. **▶ Live demo: [mecenai.onrender.com](https://mecenai.onrender.com)** · MCP endpoint: `https://mecenai.onrender.com/mcp/` (the free instance sleeps when idle — the first request may take up to a minute)
  6. **Status:** v0.4 (November 2026) — concierge on a real LLM (Groq), model evaluation, an MCP server for other AI agents, and a public demo. This README describes where the product is going; each capability is marked with its current state.
  7. ## Why
  8. An artist's or gallery's website is often a beautiful catalogue that cannot answer a question. Collectors write at night, in other languages, and wait days for a reply. Mecenai turns the website into a conversation that leads to a viewing and a sale — in the artist's own voice, under the artist's own rules.
  9. ## What it will do
  10. | Capability | How | State |
  11. |---|---|---|
  12. | Knows every artwork: story, technique, size, series, intent | retrieval over the catalogue (RAG, vector search) | planned |
  13. | Speaks in the artist's voice, in the visitor's language (English, Arabic, Russian, …) | LLM with artist-approved instructions | **v0.2** — gpt-oss-120b via Groq |
  14. | Quotes prices **only** by rules the artist approved — never negotiates or invents | tool calls with hard rules + tests | **v0.1** — scripted model |
  15. | Understands who is asking (collector, curator, press) and guides to a viewing | agent with state | planned |
  16. | Hands off to a human in Telegram / WhatsApp | webhooks, n8n | planned |
  17. | Exposes the catalogue to other AI agents | MCP server | **v0.3** — tools, resource, prompt |
  18. | Optional certificate of authenticity on TON | TON, IPFS | planned |
  19. | Insights for the artist: what collectors ask about | analytics | planned |
  20. | Measured quality: no invented prices or facts | automated evaluations | **started** — [model comparison](docs/evals/2026-11-20-model-comparison.md) |
  21. | One engine, many clients — each with its own catalogue and voice | multi-tenant service, Docker, cloud | planned |
  22. ## Run it (v0.1)
  23. ```bash
  24. python -m venv .venv && source .venv/bin/activate # Windows: .venv\Scripts\activate
  25. pip install -r requirements.txt
  26. uvicorn app.main:app --reload
  27. ```
  28. ```bash
  29. curl localhost:8000/health
  30. curl -X POST localhost:8000/chat -H "content-type: application/json" \
  31. -d '{"message": "What do you have from the Quiet Harbors series?"}'
  32. ```
  33. v0.1 runs offline on a scripted model that speaks the Messages API format, so it needs no API key. The catalogue belongs to a fictional artist.
  34. ```
  35. app/
  36. main.py FastAPI: GET /health, POST /chat
  37. concierge.py agent loop: model asks for a tool -> code runs it -> model continues
  38. tools.py find_artworks, get_price_rule (artist-approved wording only)
  39. models.py scripted model for tests + any OpenAI-compatible LLM (Groq)
  40. catalog.json fictional demo catalogue
  41. mcp_server.py MCP server: tools, resource, prompt
  42. scripts/ model comparison
  43. docs/evals/ evaluation results
  44. tests/ API, price rules, hand-off, model adapter, MCP server
  45. ```
  46. ### MCP server
  47. Any MCP client (Claude Desktop, IDEs, other agents) can use the catalogue:
  48. | Kind | Name | What |
  49. |---|---|---|
  50. | tool | `find_artworks` | search by series or title, facts only |
  51. | tool | `get_price_rule` | the only way to get a price — artist-approved wording |
  52. | tool | `request_viewing` | ask the artist for a viewing (validated, the artist replies personally) |
  53. | resource | `catalog://artist` | the whole catalogue without prices |
  54. | prompt | `recommend_artwork` | pick works for a collector's wishes |
  55. Remote (any client that supports Streamable HTTP): `https://mecenai.onrender.com/mcp/`
  56. Local, Claude Desktop (`claude_desktop_config.json`):
  57. ```json
  58. {"mcpServers": {"mecenai": {"command": "python", "args": ["/path/to/mecenai/mcp_server.py"]}}}
  59. ```
  60. Real model: copy `.env.example` to `.env`, set `MECENAI_MODEL=groq` and `GROQ_API_KEY`.
  61. Tests: `pytest -q` — they also run on every push (GitHub Actions).
  62. ## Roadmap
  63. | Version | What |
  64. |---|---|
  65. | v0.1 | service skeleton, tool calls, price rules, tests — **done** |
  66. | v0.2 | real LLM backend behind the same interface; model comparison — **done** |
  67. | v0.3 | MCP server: the catalogue for other AI agents — **done** |
  68. | v0.4 | public demo and MCP over HTTP (Render), abuse limits — **done** |
  69. | v0.5 | retrieval over artwork stories (RAG) |
  70. | v0.6 | collector profiles and hand-off to Telegram |
  71. ## Principles
  72. 1. **No invented facts or prices.** If the answer is not in the catalogue or the rules, Mecenai says it will check with the artist and hands off.
  73. 2. **Honest about being an AI** when asked.
  74. 3. **Collectors' data is minimal and private** — privacy policy (including GDPR) from day one.
  75. 4. **Every number in this README comes from a measurement.**
  76. ## Open and closed
  77. The engine, the MCP server, tests and evaluation results are open. Clients' catalogues, prices and conversations are never published; demos use a fictional artist.
  78. ## Built by
  79. [Alexey](https://github.com/MonoVarTon) — architecture, backend, automation, TON — at **369 Aether**, a studio building websites for artists and galleries worldwide.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.