# Overwing (npm · overwing-mcp)

Guardrails for LLM output: pass / fail / review verdicts with calibrated confidence in under 500 ms.

- Trust score: 71/100 (medium)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-09-25

## Components

- npm · `overwing-mcp`: 71/100 (this document), [markdown](https://verifymcp.io/servers/ai-overwing-mcp/overwing-mcp.md), [page](https://verifymcp.io/servers/ai-overwing-mcp/overwing-mcp)

## Channel facts

- Registry: `npm`
- Package: `overwing-mcp`
- Version: `0.5.0`
- 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 2026-09-25.

- **Supply Chain Security**: 98/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - No install/post-install scripts declared.
  - 31 of 95 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).
  - Actively maintained (last published 0 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 83/100
  - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (good).
  - Tool/resource definitions use about 1141 tokens (~81/item across 14 items; 13 tools + 1 resources), lean.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 7/100
  - Stability observed for 2 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 79/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 36% 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 13 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 14 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 Overwing MCP server?

Overwing runs locally as an npm package, launched with npx -y overwing-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 ai-overwing-mcp -- npx -y overwing-mcp
```

### Cursor

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

### VS Code

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

### Codex

```bash
codex mcp add ai-overwing-mcp -- npx -y overwing-mcp
```

### opencode

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

### OpenClaw

```bash
openclaw mcp add ai-overwing-mcp --command npx --arg -y --arg overwing-mcp
```

### Hermes

```yaml
mcp_servers:
  ai-overwing-mcp:
    command: "npx"
    args: ["-y", "overwing-mcp"]
```

### Netclaw

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

### Vellum

```bash
assistant mcp add ai-overwing-mcp -t stdio -c npx -a -y overwing-mcp
```

### Other

```json
{
  "mcpServers": {
    "ai-overwing-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "overwing-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.

### 2026-09-25 (score 71, +1)

- [security regression] Stability: 0.03 → unverified
- [security regression] Tool safety: pass → unverified
- [functional regression] Schema quality: 69 → 81
- [functional regression] Capabilities: pass → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Schema quality: 100 → unverified
- [functional improvement] Tool coverage: 26% → 36%
- [functional] Package version: 0.4.0 → 0.5.0
- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-09-24 (score 70, +16)

- [security regression] Tool safety: pass → unverified
- [security improvement] Malware scan: unverified → pass
- [security] Stability: Stability not yet verified: we do not have a sandbox capture of the MCP schema this version of the package serves yet.
- [functional regression] Schema quality: 50 → 69
- [functional regression] Capabilities: pass → unverified
- [functional regression] Tool coverage: 100 → unverified
- [functional regression] Schema quality: 100 → unverified
- [functional improvement] Tool coverage: 18% → 26%
- [functional improvement] Stability: unverified → 0.03
- [functional] Package version: 0.3.0 → 0.4.0

### 2026-09-23 (score 54)

First indexed and scored.

## MCP tools (13)

### `evaluate` (~256 tokens)

Evaluate text

Score any text (typically an LLM's output) against an Overwing rule set. Returns an aggregate verdict of pass, fail, or review, a recommended_action (block, redact, review, or allow), and per-rule answers with probability, confidence, and the rule's action. Act on recommended_action: block means do not send, redact means remove the flagged content and resend, review means ask a human or a slower model, allow means proceed. Prebuilt sets: 'content-safety' (toxicity, PII, self-harm, sexual content, severity) and 'outbound-message', which also takes a context object (recipient, channel, owns_contact_info) so PII that the recipient already owns is not flagged.

Input parameters:

- `context` (object): Facts the rules may reference: recipient, channel, whether you own the data, sender, purpose. Sent to the model alongside the text (max 8 KB). Use the 'outbound-message' rule set to have it honoured.
- `input` (string, required): The text to evaluate
- `metadata` (object): Opaque data stored with the evaluation and echoed in webhooks (max 8 KB)
- `rule_set` (string): Rule set slug

### `evaluate_batch` (~112 tokens)

Evaluate many texts

Score up to 50 texts against one rule set in a single call. Each item counts as one evaluation. Returns a summary plus per-item verdicts; items can fail independently.

Input parameters:

- `context` (object): Facts the rules may reference: recipient, channel, whether you own the data, sender, purpose. Sent to the model alongside the text (max 8 KB). Use the 'outbound-message' rule set to have it honoured.
- `items` (array, required)
- `rule_set` (string)

### `list_rule_sets` (~31 tokens)

List rule sets

List the prebuilt and custom rule sets available to this organization.

Input parameters:

- `include_inactive` (boolean)

### `get_rule_set` (~41 tokens)

Get rule set

Fetch a rule set with its full rule definitions. Use 'content-safety' as a worked example when writing your own.

Input parameters:

- `slug` (string, required)

### `create_rule_set` (~121 tokens)

Create rule set

Create a custom rule set with 1 to 25 rules. Each rule is a choice (pick one option), score (position on an ordered scale), or noul (yes/no) question with a fail condition, optional review threshold, weight, and action (block, redact, or review) that callers should take when it fails. Rule instructions may reference `context.*` fields that callers pass with each evaluation.

Input parameters:

- `description` (string)
- `name` (string, required)
- `rules` (array, required)
- `slug` (string, required)

### `get_evaluation` (~32 tokens)

Get evaluation

Fetch a stored evaluation by id, including the original input and per-rule results.

Input parameters:

- `id` (string, required)

### `list_evaluations` (~60 tokens)

List evaluations

List recent evaluations, newest first, with optional verdict and rule set filters. Use next_cursor to page.

Input parameters:

- `cursor` (string)
- `limit` (integer)
- `rule_set` (string)
- `verdict` (string)

### `get_usage` (~27 tokens)

Get usage

Daily usage, remaining quota for today, and plan limits.

Input parameters:

- `days` (integer)

### `whoami` (~26 tokens)

Who am I

Identify the organization and plan behind the configured API key, and whether billing is set up.

### `list_plans` (~29 tokens)

List plans

Public plan catalog: prices, daily limits, and per-minute burst limits. No API key needed.

### `atlas_lookup` (~142 tokens)

Identify a user agent (Overwing Atlas)

Say what a User-Agent string claims to be and whether the claim can be trusted, from the Overwing Atlas registry of 241 AI crawlers, fetchers and browser agents. Returns the claimed agent, operator, purpose class, verification method (Web Bot Auth signature, user-agent string only, or unattributable) and a trust note. Metered per day by Atlas tier: free 100, Pro 10,000, Team 100,000. Use it when deciding whether to serve, block, or pay-gate a request, or to understand who is hitting a site.

Input parameters:

- `user_agent` (string, required): The User-Agent header value to identify

### `atlas_agents` (~166 tokens)

Search the agent registry (Overwing Atlas)

Browse or search Overwing Atlas, the registry of AI crawlers, fetchers and browser agents: name, operator, user-agent tokens, Web Bot Auth key directory, robots.txt behaviour, and (with Atlas Pro or Team) purpose class, verification, evasion flags and traffic shares. Filter by free text, purpose (training, search, browser, coding), operator, or verification.

Input parameters:

- `limit` (integer)
- `operator` (string)
- `purpose` (string): Purpose substring, e.g. training, search, browser, fetcher, coding
- `q` (string): Free text over name, operator, user agents, description
- `verification` (string): e.g. 'Web Bot Auth', 'spoofable', 'Unattributable'

### `atlas_summary` (~72 tokens)

Agent traffic and spending summary (Overwing Atlas)

Public numbers from Overwing Atlas: registry counts by purpose and verification, published browser-agent traffic shares, sector field-scan headlines (e.g. how many OSINT sites carry AI-crawler rules or any agent-payable surface), and the summary of the Agent Consumers report on what agents actually spend. No key needed.

## Diagnostics

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

## Score history

- 2026-09-25: 71
- 2026-09-24: 70
- 2026-09-23: 54

## Common questions

### What is the Overwing MCP server?

Overwing is an MCP server listed in the public MCP registry as ai.overwing/mcp. Guardrails for LLM output: pass / fail / review verdicts with calibrated confidence in under 500 ms. This page covers its npm package (overwing-mcp).

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

Overwing scores 71 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 25 September 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 Overwing MCP server expose?

Overwing exposes 13 tools: evaluate, evaluate_batch, list_rule_sets, get_rule_set, create_rule_set, and 8 more. Their descriptions and schemas cost roughly 1,115 tokens of context every time the server is loaded.

### Is the Overwing MCP server still maintained?

Overwing is still listed as active in the MCP registry. We last reached this channel on 25 September 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 Overwing MCP server under?

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

## Links

- npm package: https://www.npmjs.com/package/overwing-mcp
- Socket report: https://socket.dev/npm/package/overwing-mcp
- Repository: https://github.com/frod27/overwing-mcp
- Website: https://overwing.ai/
- Changelog RSS feed: https://verifymcp.io/servers/ai-overwing-mcp/overwing-mcp.xml
- Changelog JSON feed: https://verifymcp.io/servers/ai-overwing-mcp/overwing-mcp.json
- HTML version of this page: https://verifymcp.io/servers/ai-overwing-mcp/overwing-mcp