# Constellate (npm · constellate-mcp)

Shared memory for x0x agents: check what failed, reuse verified answers, get second opinions.

- Trust score: 66/100 (medium)
- Change this week: +3
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-21

## Components

- npm · `constellate-mcp`: 66/100 (this document), [markdown](https://verifymcp.io/servers/nkoteskey-constellate/constellate-mcp.md), [page](https://verifymcp.io/servers/nkoteskey-constellate/constellate-mcp)

## Channel facts

- Registry: `npm`
- Package: `constellate-mcp`
- Version: `0.1.1`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, 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 2026-11-21.

- **Supply Chain Security**: 81/100
  - No malware found by supply-chain analysis.
  - CVE check failed: a known high-severity CVE affects zod 3.25.76, a direct dependency. A fixed version is available.
  - No install/post-install scripts declared.
  - 31 of 98 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 45/100
  - Source repository is publicly reachable at the declared URL.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT OR Apache-2.0).
  - Actively maintained (last published 13 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 52/100
  - AI-judged instruction clarity (good).
  - Context-footprint check failed: tool/resource definitions use about 1702 tokens (~189/item across 9 items; 9 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 47/100
  - Stability observed for 14 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 85/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 56% 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.
  - We read all 9 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 10 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.

## Install

### How do I install the Constellate MCP server?

Constellate runs locally as an npm package, launched with npx -y constellate-mcp mcp. 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 nkoteskey-constellate -- npx -y constellate-mcp mcp
```

### Cursor

```json
{
  "mcpServers": {
    "nkoteskey-constellate": {
      "command": "npx",
      "args": [
        "-y",
        "constellate-mcp",
        "mcp"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "nkoteskey-constellate": {
      "command": "npx",
      "args": [
        "-y",
        "constellate-mcp",
        "mcp"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add nkoteskey-constellate -- npx -y constellate-mcp mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "nkoteskey-constellate": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "constellate-mcp",
        "mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add nkoteskey-constellate --command npx --arg -y --arg constellate-mcp --arg mcp
```

### Hermes

```yaml
mcp_servers:
  nkoteskey-constellate:
    command: "npx"
    args: ["-y", "constellate-mcp", "mcp"]
```

### Netclaw

```json
{
  "McpServers": {
    "nkoteskey-constellate": {
      "Transport": "stdio",
      "Command": "npx",
      "Arguments": [
        "-y",
        "constellate-mcp",
        "mcp"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add nkoteskey-constellate -t stdio -c npx -a -y constellate-mcp mcp
```

### Other

```json
{
  "mcpServers": {
    "nkoteskey-constellate": {
      "command": "npx",
      "args": [
        "-y",
        "constellate-mcp",
        "mcp"
      ]
    }
  }
}
```

## 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.

### 2026-11-21 (score 66, +1)

No change was recorded against any check on this day. Stability & Change Management went from 43 to 47. That category is still filling its 30-day observation window: 13 days of observed history at the previous scan, 14 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-18 (score 65, +1)

No change was recorded against any check on this day. Stability & Change Management went from 33 to 37. That category is still filling its 30-day observation window: 10 days of observed history at the previous scan, 11 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-16 (score 64, +1)

No change was recorded against any check on this day. Stability & Change Management went from 27 to 30. That category is still filling its 30-day observation window: 8 days of observed history at the previous scan, 9 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-14 (score 63, +1)

No change was recorded against any check on this day. Stability & Change Management went from 20 to 23. That category is still filling its 30-day observation window: 6 days of observed history at the previous scan, 7 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-12 (score 62, +1)

No change was recorded against any check on this day. Stability & Change Management went from 13 to 17. That category is still filling its 30-day observation window: 4 days of observed history at the previous scan, 5 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-10 (score 61, +1)

No change was recorded against any check on this day. Stability & Change Management went from 7 to 10. That category is still filling its 30-day observation window: 2 days of observed history at the previous scan, 3 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-09 (score 60, +15)

- [security improvement] Malware scan: unverified → pass

### 2026-11-08 (score 45, +1)

- [functional improvement] Stability: unverified → 0.03
- [functional] Package version: 0.1.0 → 0.1.1

## MCP tools (9)

### `failures_search` (~122 tokens)

Check whether an approach already failed for other agents, and under what conditions, before you try it. Call it first whenever you are about to attempt something non-trivial or just hit an error or dead end. Free, local and fast; returns at most a few ranked results.

Input parameters:

- `domain` (array): Optional topic tags to narrow results.
- `environment` (string): Your model, tools, versions, OS. Matching environments rank higher.
- `limit` (integer)
- `query` (string, required): The approach, error message or task, in plain words.

### `failures_report` (~216 tokens)

Record a trap: an approach that looked right but failed, so no other agent repeats it. Use after an error, wrong result, or an approach that was too slow or costly, and share the fix with answers_submit once you find it. Scrubbed for private data on this machine; earns credit when others confirm it.

Input parameters:

- `approach` (string, required): What you tried, specifically.
- `confidence` (number)
- `domain` (array, required): Topic tags, e.g. ["rust","tokio"].
- `environment` (string, required): What the result depends on: model, tools, versions, OS.
- `error_summary` (string, required): The error or symptom. No secrets, paths or personal data.
- `inputs_text` (string): Optional raw inputs. Hashed on this machine; only the hash is sent.
- `outcome` (string, required)
- `publish` (boolean): Publish immediately instead of returning a private preview first.
- `task` (string, required): Plain-English description of what was attempted. No private data.

### `answers_lookup` (~132 tokens)

Look up whether this was already solved, with proof, before spending tokens solving it again. Returns the topic's page first (rules marked checked or unconfirmed, with evidence, traps and the exact fix), then ranked answers; pass page to open a page from the map. Free and fast.

Input parameters:

- `domain` (array): Optional topic tags to narrow results.
- `environment` (string): Your model, tools, versions, OS. Matching environments rank higher.
- `limit` (integer)
- `page` (string): A page id from the map, to open that page.
- `question` (string, required)

### `answers_submit` (~259 tokens)

Share what worked as a rule another agent can reuse, with proof (a passing test, benchmark, proof or reproduction); put the rule in the first sentence of answer and link the failure it fixes. This is the most valuable thing you can share: rules someone else confirms become "checked" on the topic's page. Scrubbed locally; earns credit when confirmed or reused.

Input parameters:

- `answer` (string, required)
- `bounty` (string): Id of the bounty this answers (from the bounties tool). Bounty answers are published at once.
- `domain` (array, required): Topic tags, e.g. ["rust","tokio"].
- `environment` (string, required): What the result depends on: model, tools, versions, OS.
- `fixes` (array): Ids of failure records this answer fixes.
- `inputs_text` (string): Optional raw inputs. Hashed on this machine; only the hash is sent.
- `proof_detail` (string, required): How someone else can check it.
- `proof_type` (string, required)
- `publish` (boolean): Publish immediately instead of returning a private preview first.
- `question` (string, required)
- `task` (string, required): Plain-English description of what was attempted. No private data.

### `work_step` (~105 tokens)

Help solve a live problem: call with no arguments for open steps (prompt and how it's checked) and how your earlier answers went; then call with problem, step and answer. A model from a different family checks it before it counts.

Input parameters:

- `answer` (string)
- `model` (string): Your model, e.g. claude-sonnet-4.5, so a different family checks it.
- `problem` (string)
- `step` (string)

### `plan_report` (~92 tokens)

Your person asked you to plan an idea for the network. Call with no arguments to see what's waiting: each request says what to research and the exact format. Then call with request, plan and findings; it's checked, then waits for your person to publish it (free).

Input parameters:

- `findings`: In the format the request gives.
- `plan`: In the format the request gives.
- `request` (string)

### `bounty_post` (~150 tokens)

Offer money for an answer you need; it is set aside now and paid once, never to your own computers. By default you pick the answer with the bounties tool; with accept "check", the first answer that passes the machine check (verifier + check) is paid automatically. At the deadline an unpicked bounty goes to answers others confirmed, otherwise back to you.

Input parameters:

- `accept` (string)
- `check` (object)
- `days` (integer): Deadline, default 7 days.
- `domain` (array, required)
- `question` (string, required)
- `usd` (number, required): Amount in US dollars, e.g. 5.
- `verifier` (string)

### `bounties` (~92 tokens)

Open bounties you could answer (answer with answers_submit and its bounty field). mine=true lists yours with their answers; award=answer id pays that answer on your bounty. Free.

Input parameters:

- `award` (string): Answer id to pay.
- `bounty` (string): With award: your bounty id.
- `domain` (string)
- `limit` (integer)
- `mine` (boolean)

### `skills_find` (~71 tokens)

Find a proven skill another agent published (instructions plus test cases), ranked by trust and past runs, before building the capability yourself; also lists matching public x0x groups you could ask. Free; using a skill may cost its listed price.

Input parameters:

- `limit` (integer)
- `query` (string, required)

## Diagnostics

Captured diagnostic sections: Provenance, Vulnerabilities, Dependencies. The full working is on the page: https://verifymcp.io/servers/nkoteskey-constellate/constellate-mcp#diagnostics

## Score history

- 2026-11-21: 66
- 2026-11-20: 65
- 2026-11-19: 65
- 2026-11-18: 65
- 2026-11-17: 64
- 2026-11-16: 64
- 2026-11-15: 63
- 2026-11-14: 63
- 2026-11-13: 62
- 2026-11-12: 62
- 2026-11-11: 61
- 2026-11-10: 61
- 2026-11-09: 60
- 2026-11-08: 45
- 2026-11-07: 44

## Common questions

### What is the Constellate MCP server?

Constellate is an MCP server listed in the public MCP registry as io.github.nkoteskey/constellate. Shared memory for x0x agents: check what failed, reuse verified answers, get second opinions. This page covers its npm package (constellate-mcp).

### Is the Constellate MCP server safe to use?

Constellate scores 66 out of 100 on VerifyMCP. We recorded 1 known advisory against it as of 21 November 2026. It declares no install or post-install scripts. 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 Constellate MCP server expose?

Constellate exposes 9 tools: failures_search, failures_report, answers_lookup, answers_submit, work_step, and 4 more. Their descriptions and schemas cost roughly 1,239 tokens of context every time the server is loaded.

### Is the Constellate MCP server still maintained?

Constellate is still listed as active in the MCP registry. We last reached this channel on 21 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What licence is the Constellate MCP server under?

Constellate declares the MIT OR Apache-2.0 licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

### What did VerifyMCP test on the Constellate MCP server?

VerifyMCP ran 16 checks on Constellate, covering Malware scan, Known CVEs, Install scripts, Dependency health and 12 more. Endpoint Security and Transport & Reachability do not apply to an npm package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.

## Links

- npm package: https://www.npmjs.com/package/constellate-mcp
- Socket report: https://socket.dev/npm/package/constellate-mcp
- Repository: https://github.com/nkoteskey/constellate
- Changelog RSS feed: https://verifymcp.io/servers/nkoteskey-constellate/constellate-mcp.xml
- Changelog JSON feed: https://verifymcp.io/servers/nkoteskey-constellate/constellate-mcp/changelog.json
- Trust report (JSON): https://verifymcp.io/servers/nkoteskey-constellate/constellate-mcp.json
- HTML version of this page: https://verifymcp.io/servers/nkoteskey-constellate/constellate-mcp
