# PaKi Curator (remote · paki-api.elfresonero.workers.dev)

Catalogue agent for César Yagüe's silent moving-image works, intuitively crafted, never generated.

- Trust score: 75/100 (medium)
- Change this week: 0
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-10-23

## Components

- remote · `paki-api.elfresonero.workers.dev`: 75/100 (this document), [markdown](https://verifymcp.io/servers/es-cesaryague-paki-curator/paki-api.md), [page](https://verifymcp.io/servers/es-cesaryague-paki-curator/paki-api)

## Channel facts

- Endpoint: `https://paki-api.elfresonero.workers.dev/mcp`
- Transports: `streamable-http`
- Auth: `none`
- Version: `2.0.3`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-10-23.

- **Endpoint Security**: 63/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - No authorisation is required to call this server. Every tool declares its destructiveHint and none is destructive, so open access doesn't expose one.
  - HTTPS check failed: the endpoint is reachable over plaintext HTTP.
  - HSTS check failed: the Strict-Transport-Security header is absent.
  - DNSSEC check failed: this domain isn't protected by DNSSEC.
- **Transport & Reachability**: 100/100
  - Verified streamable-http transport via a live MCP handshake.
- **Schema Quality & AI Usability**: 63/100
  - AI-judged instruction clarity (good).
  - Context-footprint check failed: tool/resource definitions use about 1383 tokens (~172/item across 8 items; 8 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 100/100
  - No destabilizing schema changes in the last 30 days.
- **Tool Coverage**: 86/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 58% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - All 1 tool(s) whose name or description implies an irreversible operation declare an MCP destructiveHint annotation.
  - An AI judge read all 9 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 40/100
  - Spec-recency check failed: implements MCP spec 2025-03-26; the latest is 2026-07-28.

## Install

### How do I install the PaKi Curator MCP server?

PaKi Curator is a hosted endpoint at https://paki-api.elfresonero.workers.dev/mcp, so there is nothing to install locally. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add --transport http es-cesaryague-paki-curator 'https://paki-api.elfresonero.workers.dev/mcp'
```

### Cursor

```json
{
  "mcpServers": {
    "es-cesaryague-paki-curator": {
      "url": "https://paki-api.elfresonero.workers.dev/mcp"
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "es-cesaryague-paki-curator": {
      "type": "http",
      "url": "https://paki-api.elfresonero.workers.dev/mcp"
    }
  }
}
```

### Codex

```toml
[mcp_servers.es-cesaryague-paki-curator]
url = "https://paki-api.elfresonero.workers.dev/mcp"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "es-cesaryague-paki-curator": {
      "type": "remote",
      "url": "https://paki-api.elfresonero.workers.dev/mcp",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add es-cesaryague-paki-curator --url 'https://paki-api.elfresonero.workers.dev/mcp' --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  es-cesaryague-paki-curator:
    url: "https://paki-api.elfresonero.workers.dev/mcp"
```

### Netclaw

```json
{
  "McpServers": {
    "es-cesaryague-paki-curator": {
      "Transport": "http",
      "Url": "https://paki-api.elfresonero.workers.dev/mcp"
    }
  }
}
```

### Vellum

```bash
assistant mcp add es-cesaryague-paki-curator -t streamable-http -u 'https://paki-api.elfresonero.workers.dev/mcp'
```

### Other

```json
{
  "mcpServers": {
    "es-cesaryague-paki-curator": {
      "type": "http",
      "url": "https://paki-api.elfresonero.workers.dev/mcp"
    }
  }
}
```

The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs.

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-10-17 (score 75, 0)

- [security] The server rewrote its instructions, which are the text every model session reads
- [security] Tool “search_artworks” rewrote its description, which is the text the model reads
- [cosmetic] Tool “catalog_overview” changed its title: Catalog overview → Catalogue overview

### 2026-10-10 (score 75, 0)

- [security] Stability: 0.97 → pass
- [security] Tool “request_inquiry” rewrote its description, which is the text the model reads
- [functional regression] Schema quality: 1205 → 1383
- [functional] Destructive annotations: pass → 100
- [functional] New tool “confirm_inquiry”

### 2026-10-06 (score 75, +1)

No change was recorded against any check on this day. Stability & Change Management went from 90 to 97. That category is still filling its 30-day observation window: 27 days of observed history at the previous scan, 29 at this one. The score rises as the window fills, whether or not the server changes. Other categories moved too: Schema Quality & AI Usability fell 1.

### 2026-10-04 (score 74, +1)

No change was recorded against any check on this day. Stability & Change Management went from 87 to 90. That category is still filling its 30-day observation window: 26 days of observed history at the previous scan, 27 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-03 (score 73, 0)

- [security] Tool “recommend_for_space” rewrote its description, which is the text the model reads
- [security] Tool “search_artworks” rewrote its description, which is the text the model reads
- [cosmetic] “recommend_for_space” added an optional parameter “resolution”

### 2026-10-02 (score 73, 0)

- [security] Tool “search_artworks” rewrote its description, which is the text the model reads
- [cosmetic] “recommend_for_space” added an optional parameter “collection”

### 2026-10-01 (score 73, +1)

No change was recorded against any check on this day. Stability & Change Management went from 77 to 80. That category is still filling its 30-day observation window: 23 days of observed history at the previous scan, 24 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-09-30 (score 72, 0)

- [security] The server rewrote its instructions, which are the text every model session reads
- [security] Tool “search_artworks” rewrote its description, which is the text the model reads
- [functional regression] Schema quality: 139 → 166
- [functional improvement] Tool coverage: 47% → 54%
- [cosmetic] “search_artworks” added an optional parameter “max_seconds”
- [cosmetic] “search_artworks” added an optional parameter “year”
- [cosmetic] “search_artworks” added an optional parameter “rotation”
- [cosmetic] “search_artworks” added an optional parameter “offset”
- [cosmetic] “search_artworks” added an optional parameter “min_seconds”
- [cosmetic] “search_artworks” reworded the description of “resolution”

## MCP tools (8)

### `search_artworks` (~361 tokens)

Search artworks

Search the catalogue of 297 moving-image works by César Yagüe by free text, collection, resolution, orientation (including ultrawide), colour measured from every frame (a basic colour counts its whole family), price, year, loop length, rotation and availability. A short page brings each work's curatorial capsule. Asking about flashes or epilepsy returns the works with measured abrupt light events.

Input parameters:

- `availability` (string)
- `collection` (string): Collection name
- `colour` (string): Dominant colour by name (gold, teal, crimson…) or hex. Measured from every frame.
- `limit` (integer)
- `max_price` (number)
- `max_seconds` (number): Longest loop length, in seconds
- `min_price` (number)
- `min_seconds` (number): Shortest loop length, in seconds
- `offset` (integer): Skip this many results, to read the next page
- `order` (string)
- `orientation` (string): ultrawide = at least twice as wide as tall (21:9 and wider); landscape includes it
- `query` (string): Free text across titles, descriptions and curatorial notes
- `resolution` (string): The resolution label each work shows, matched exactly: HD, 4K, 5K, 8K, 10K, 12K, 16K; square works carry their own, such as 2K or 3K. An unknown label returns the list in use.
- `rotation` (string): Whether the work must keep its orientation or may be rotated 90 degrees
- `sort_by` (string)
- `year` (integer): Year the work was created

### `get_artwork` (~86 tokens)

Get artwork

Retrieve one artwork in full: technical specification, availability, price, the physics measured from its master, and, when available, the curatorial capsule. Accepts the identifier or the title — and when a title belongs to more than one work, it returns them all rather than guessing.

Input parameters:

- `name` (string, required): Artwork identifier, or its name. An ambiguous name returns every candidate.

### `recommend_for_space` (~280 tokens)

Recommend for space

Describe a space and receive the works that fit it, each with the measurement behind the suggestion and its curatorial capsule. Room light, viewing distance, screen size, attention and movement are measured; collection, resolution, orientation, colour and budget are hard limits, and a work that may be rotated counts for the other orientation. If nothing fits, it says which requirement leaves nothing instead of offering the nearest thing.

Input parameters:

- `attention` (string): Should it accompany, or hold the eye?
- `budget_max` (number)
- `collection` (string): Recommend only from this collection
- `colour` (string)
- `count` (integer)
- `distance_m` (number): Metres from viewer to screen
- `movement` (string): Very still, or full of life? Measured rhythm, on the catalogue scale
- `orientation` (string): ultrawide = at least twice as wide as tall (21:9 and wider); landscape includes it
- `resolution` (string): Only works native to this resolution label (HD, 4K, 8K, 16K…; square works carry their own, such as 2K)
- `room_light` (string)
- `screen_m` (number): Screen diagonal in metres
- `space_description` (string, required): The room in your own words

### `browse_collections` (~64 tokens)

Browse collections

List the 15 collections plus Singles, each with the number of works it holds. Ask for one by name to get its works in full. Each collection includes its curatorial capsule alongside the works it contains.

Input parameters:

- `collection` (string): Omit to list them all

### `catalog_overview` (~40 tokens)

Catalogue overview

A high-level view of the whole catalogue: totals, collections, resolutions, price range, what is available, and how much of it has been measured.

### `rank_artworks` (~158 tokens)

Rank artworks

Order the catalogue by one measured axis — rhythm of change, luminance, peak brightness, colour travel, empty space, and the rest — from lowest to highest or the reverse, and read where each work sits on the catalogue's own scale. Use it for questions like "the stillest work you have" or "the brightest, in this collection". Every number comes from the master, frame by frame; nothing here is described.

Input parameters:

- `availability` (string)
- `collection` (string): Restrict to one collection
- `count` (integer)
- `metric` (string, required): The measured axis to order by
- `order` (string): asc = lowest first (the stillest, the darkest); desc = highest first

### `request_inquiry` (~172 tokens)

Request inquiry

Prepare an enquiry to the studio — acquisition, exhibition or press — and see exactly what would be sent. Nothing is sent at this step: the draft is held for two days, seen by nobody, and deleted unless it is confirmed. Show the details to the person named, and only if they agree to be contacted, send it with confirm_inquiry and the code this returns. Requires a way to reply and a description of the space or project.

Input parameters:

- `agent_name` (string): The agent sending this on behalf of a client
- `artworks_of_interest` (array)
- `budget_range` (string)
- `contact_email` (string, required)
- `contact_name` (string, required)
- `inquiry_type` (string)
- `message` (string)
- `space_description` (string, required)

### `confirm_inquiry` (~97 tokens)

Confirm inquiry

Send an enquiry prepared with request_inquiry, once the person named has seen exactly what will be sent and agreed to be contacted. Takes the confirmation code and the same email; what is sent is exactly what was shown, and only once. It never reports success unless the message actually left.

Input parameters:

- `confirmation_code` (string, required): The code request_inquiry returned
- `contact_email` (string, required): The same email the enquiry was prepared with

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/es-cesaryague-paki-curator/paki-api#diagnostics

## Score history

- 2026-10-23: 75
- 2026-10-22: 75
- 2026-10-21: 75
- 2026-10-20: 75
- 2026-10-19: 75
- 2026-10-18: 75
- 2026-10-17: 75
- 2026-10-16: 75
- 2026-10-15: 75
- 2026-10-14: 75
- 2026-10-13: 75
- 2026-10-12: 75
- 2026-10-10: 75
- 2026-10-06: 75
- 2026-10-04: 74
- 2026-10-03: 73
- 2026-10-02: 73
- 2026-10-01: 73
- 2026-09-30: 72
- 2026-09-29: 72
- 2026-09-28: 64
- 2026-09-27: 64
- 2026-09-26: 63
- 2026-09-25: 63
- 2026-09-24: 63
- 2026-09-23: 62
- 2026-09-22: 62
- 2026-09-21: 62
- 2026-09-20: 61
- 2026-09-19: 61

## Common questions

### What is the PaKi Curator MCP server?

PaKi Curator is an MCP server listed in the public MCP registry as es.cesaryague/paki-curator. Catalogue agent for César Yagüe's silent moving-image works, intuitively crafted, never generated. This page covers its hosted endpoint (https://paki-api.elfresonero.workers.dev/mcp).

### Is the PaKi Curator MCP server safe to use?

PaKi Curator scores 75 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the PaKi Curator MCP server expose?

PaKi Curator exposes 8 tools: search_artworks, get_artwork, recommend_for_space, browse_collections, catalog_overview, and 3 more. Their descriptions and schemas cost roughly 1,258 tokens of context every time the server is loaded.

### Does the PaKi Curator MCP server require authentication?

No. We connected to PaKi Curator without credentials and it answered, so anything it exposes is reachable by anyone who knows the address.

### Is the PaKi Curator MCP server still maintained?

PaKi Curator is still listed as active in the MCP registry. We last reached this channel on 23 October 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

## Links

- Remote endpoint: https://paki-api.elfresonero.workers.dev/mcp
- Website: https://cesaryague.es/
- Changelog RSS feed: https://verifymcp.io/servers/es-cesaryague-paki-curator/paki-api.xml
- Changelog JSON feed: https://verifymcp.io/servers/es-cesaryague-paki-curator/paki-api.json
- HTML version of this page: https://verifymcp.io/servers/es-cesaryague-paki-curator/paki-api