Search DevTools

Jump to any tool or page

Overwing

Guardrails for LLM output: pass / fail / review verdicts and one recommended action. Hosted or npm.

frod270 stars0 forksAI & Agents
View source

Install

Terminal

$npx -y overwing-mcp

mcp_config.json

{
  "mcpServers": {
    "ai-overwing-mcp": {
      "env": {
        "OVERWING_API_KEY": "${OVERWING_API_KEY}",
        "OVERWING_BASE_URL": "${OVERWING_BASE_URL}",
        "OVERWING_AGENT_KEY": "${OVERWING_AGENT_KEY}"
      },
      "args": [
        "-y",
        "overwing-mcp"
      ],
      "command": "npx"
    }
  }
}

Documentation

Guardrails for LLM output, as MCP tools. Score any text for safety and quality. Get pass / fail / review verdicts with calibrated confidence, usually in under 500 ms.

简体中文 · 日本語 · 한국어

Overwing is an API that checks what your model said before it ships. This package exposes it to any MCP-capable agent: Claude Desktop, Claude Code, Cursor, Windsurf, VS Code, OpenAI's Agents SDK, and anything else that speaks the Model Context Protocol.

  • Real verdicts, not vibes. Every rule returns a typed answer, a probability, and a confidence. fail means a rule matched; review means it was unsure; pass means neither.
  • Prebuilt content-safety and outbound-message rule sets: toxicity, personal data, confidential leaks, self-harm, sexual content, severity. Or write your own rules in plain language.
  • Built for agents. Sign up, pay, evaluate, rotate keys, and cancel, all as JSON. No CAPTCHA, no browser required. See overwing.ai/llms.txt.

Try it without installing anything: paste text into the console at overwing.ai.

Hosted, no install

The same 34 tools are served at https://overwing.ai/mcp (Streamable HTTP). Point any client that takes a remote MCP URL at it:

claude mcp add --transport http overwing https://overwing.ai/mcp

No key is needed for evaluate, atlas_lookup, atlas_summary, list_plans, create_account and the three beacon_ tools. For the rest, send Authorization: Bearer ow_live_... on the connection.

Install

It works with no API key: evaluate and atlas_lookup each run 10 times a day free, and text sent without a key is not stored. For more, get a key at overwing.ai/login, or let your agent make its own account with the create_account tool. That account has no email: nothing is sent to anyone, and the key is returned once. The same call from a shell:

curl -X POST https://overwing.ai/api/v1/signup

Put the key in OVERWING_API_KEY. An account with no email starts at 50 evaluations a day; proving a domain (prove_domain, verify_domain) raises it to the normal free limits and makes a lost key recoverable.

Claude Code

claude mcp add overwing -e OVERWING_API_KEY=ow_live_... -- npx -y overwing-mcp

Claude Desktop, Cursor, Windsurf, VS Code (any JSON-configured client)

{
  "mcpServers": {
    "overwing": {
      "command": "npx",
      "args": ["-y", "overwing-mcp"],
      "env": { "OVERWING_API_KEY": "ow_live_..." }
    }
  }
}

Set OVERWING_BASE_URL to point at a self-hosted deployment. Requires Node 20+.

For Overwing Tower, set OVERWING_AGENT_KEY=ow_agent_... to operate as an existing agent. Without it, tower_create_agent mints a key with the organization key and keeps it in memory for the session.

evaluate, atlas_lookup, atlas_summary and list_plans work with no key at all.

The text can be in any language. It was tested on 2026-09-29 in Spanish, Portuguese, French, German, Japanese, Simplified Chinese, Korean, Arabic and Hindi: a small test, not a benchmark. Results come back in English.

Tools

ToolWhat it does
evaluateScore one text against a rule set. Returns the verdict, a recommended_action (block, redact, review, or allow), aggregate score, confidence, latency, and per-rule results with each rule's action. Takes an optional context object (recipient, channel, ownership) that context-aware rule sets such as outbound-message read, and store: false to run the check without keeping the text.
evaluate_batchScore up to 50 texts in one call, with a summary and per-item verdicts and recommended actions.
list_rule_sets · get_rule_set · create_rule_setBrowse the prebuilt set or define your own rules: yes/no questions, classifications, or scored scales.
get_evaluation · list_evaluationsRead stored results, filter by verdict or rule set, page with a cursor.
create_account · prove_domain · verify_domainAn account for the agent itself, with no email: create_account returns a key and sends nothing to anyone. Proving a domain (a DNS TXT record or a file) takes the place of the email: it raises the limits, opens full Beacon reports, and makes a lost key recoverable.
atlas_lookupOverwing Atlas: say what a User-Agent string claims to be and whether the claim can be trusted (Web Bot Auth, spoofable string, or unattributable). Works with no key, 10 a day; 100 a day with a free key.
atlas_agents · atlas_summarySearch the registry of AI crawlers, fetchers and browser agents; get traffic shares, sector field-scan headlines, and the agent-spending summary.
atlas_register_agent · atlas_verify_registration · atlas_list_registrations · atlas_withdraw_registrationAdd an agent you operate to the Atlas registry, free, so a lookup of its User-Agent names you. Register it, publish the value you are given at the operator's domain (a DNS TXT record or a file), then verify. Needs OVERWING_API_KEY.
beacon_start · beacon_report · beacon_sampleOverwing Beacon: is a site reachable by agents? Checks robots.txt, llms.txt, the MCP server card, endpoint and Registry listing, the A2A agent card and OpenAPI, and how the home page reads to a model. A check is free: beacon_start returns an id and beacon_report runs the check and returns the report. With OVERWING_API_KEY it is the full report, saved to your dashboard; with no key it is the summary (the score, the three answers and the first fix). beacon_sample is a real report in full.
tower_load_template · tower_create_agent · tower_list_agents · tower_revoke_agentOverwing Tower setup, with the organization key: load the starter order-entry workflow, then mint a scoped agent identity. The new agent key is used for the rest of the session.
tower_capabilities · tower_decide · tower_submit_action · tower_get_action · tower_compensateOverwing Tower operations, with the agent key: see which operations you may call and their input schemas, ask how a request would be ruled, submit it (executed, pending human review, or rejected), poll it, undo it.
tower_get_receipt · tower_verify_receiptsRead a signed receipt, or recompute every hash and signature in the chain.
get_usage · whoami · list_plansToday's quota, the org behind the key, and the public plan catalog.

The overwing://guide resource returns the full plain-text API guide.

Data handling and security

What this server does on your machine: it reads three environment variables (OVERWING_API_KEY, OVERWING_AGENT_KEY, OVERWING_BASE_URL) and makes HTTPS requests to the Overwing API. It reads no files, runs no commands, and talks to no other host. It has two runtime dependencies (@modelcontextprotocol/sdk, zod), and the whole server is one file: src/index.ts.

What happens to text you evaluate:

  • It is sent to the Overwing API, and from there to TypeSafe, whose Jev model produces the verdict. It is not used to train models.
  • Without a key the text is never stored.
  • With a key the text, context and verdict are stored so you can read them back with get_evaluation, until you delete them. To change that:
    • pass store: false to evaluate or evaluate_batch to keep no text or context for that call;
    • set store_inputs: false on the organization (PATCH /api/v1/org, or the dashboard) to keep none at all;
    • set retention_days on the organization to delete evaluations after that many days.
  • Traffic is encrypted in transit (TLS) and the database is encrypted at rest (AES-256).

Give your agent a restricted key. A key with scope evaluate can run checks and read rule sets, and cannot read stored evaluations, change rule sets, manage keys, webhooks or billing, or delete anything:

curl -X POST https://overwing.ai/api/v1/api-keys \
  -H "Authorization: Bearer ow_live_..." -H "Content-Type: application/json" \
  -d '{"name":"agent","scope":"evaluate"}'

Overwing is a young service and has not had an independent security audit. The full picture, including subprocessors and what we have not done, is at overwing.ai/security; the privacy policy is at overwing.ai/privacy. To report a vulnerability, see SECURITY.md.

Example

Ask your agent:

Check this reply before I send it: "Reach me at dana@example.com or 555-0142 to sort out the refund."

It calls evaluate and gets back:

Verdict: FAIL  score=0.82  confidence=0.97  251ms
  toxicity: pass (answer="safe", confidence=1)
  pii_detected: fail (answer=true, confidence=0.98)
  self_harm: pass (answer=false, confidence=1)
  sexual_content: pass (answer="none", confidence=1)
  severity: pass (answer=0.03, confidence=0.97)

How verdicts work

Each rule has a fail condition, an optional review threshold, and a weight.

  • fail: a rule's fail condition matched. Block it, redact it, or regenerate.
  • review: nothing failed, but a rule's confidence was below its threshold. Route to a person or a slower model.
  • pass: everything else.

Every rule also carries an action (block, redact, or review), and the response rolls them up into one recommended_action: block beats redact beats review beats allow. Branch on that field. Pass a context object (recipient, channel, owns_contact_info) with the outbound-message rule set and its rules read it, so a customer's own phone number in a reply to that customer is not flagged.

aggregate_score is 0 to 1 (pass = 1, review = 0.5, fail = 0 per rule, weighted). confidence is the minimum across rules.

Pricing

Free: 250 evaluations a day. Paid plans from $29/month. Every plan includes every endpoint, custom rule sets, webhooks, and the dashboard. Full details at overwing.ai/#pricing or GET https://overwing.ai/api/v1/plans.

Links

Development

npm install
npm run build
OVERWING_API_KEY=ow_live_... node dist/index.js

MIT © Overwing. Verdicts are produced by TypeSafe's Jev System One model; Overwing is not affiliated with TypeSafe.

Sourced from the repository README.

More in AI & Agents