Redacts personal data and secrets from text prompts before they are sent to an AI model.
Also known as deliverd-dev/deliverd-sdk, deliverd-screen, @deliverd/screen
- Stars
- 1
- Tools
- —
- Downloads, 30 days
- 612
- Latest version
- 0.3.0
- Last commit
- Nov 15, 2026
- License
- MIT
What is Deliverd Screen MCP?
Deliverd Screen redacts personal data and secrets from text prompts before they are sent to an AI model.
Summary generated from the project's README
About Deliverd Screen MCP server
Deliverd Screen provides a local privacy layer that identifies and redacts sensitive information from text prompts before they reach an AI model. It is designed for developers who need to prevent the accidental transmission of email addresses, IBANs, social security numbers, API keys, and other confidential data to external LLMs. The tool operates locally without requiring an account or network calls, ensuring that text stays within your process unless you explicitly enable the Microsoft Purview integration for organizational data loss prevention.
Users can customize the redaction process by plugging in external models for name detection, such as Hugging Face token classifiers or Presidio. The system uses a reversible placeholder method, replacing sensitive values with unique identifiers like [EMAIL_1] and allowing you to restore the original data once the model provides an answer. It supports various integration methods including a Vercel AI SDK middleware, a command line interface for CI/CD pipelines, and an MCP server for use with agents in Claude Code or Cursor via the screen_text tool.
What you can do with it
- 1Redact sensitive prompt dataA developer can use the screen_text tool to ensure an AI agent never sees private identifiers in a prompt.
- 2Automate privacy checks in CIYou can run automated checks on evaluation files to fail a build if sensitive data is detected.
- 3Integrate custom name detectionDevelopers can connect a local Presidio instance to identify people's names within text.
- 4Enforce organizational DLP policiesOrganizations can apply existing Microsoft Purview rules to AI prompts by connecting to their own Microsoft tenant.
Written from the project's README. Every use case is checked against a quote from it.
How to install Deliverd Screen MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall Deliverd Screen MCP in Claude Code
$ claude mcp add deliverd-screen -- npx -y @deliverd/screenInstall Deliverd Screen MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"deliverd-screen": {
"command": "npx",
"args": [
"-y",
"@deliverd/screen"
]
}
}
}Install Deliverd Screen MCP in VS Code
.vscode/mcp.json
{
"servers": {
"deliverd-screen": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@deliverd/screen"
]
}
}
}Install Deliverd Screen MCP in VS Code CLI
$ code --add-mcp '{"name":"deliverd-screen","type":"stdio","command":"npx","args":["-y","@deliverd/screen"]}'Install Deliverd Screen MCP in Claude Desktop
claude_desktop_config.json
{
"mcpServers": {
"deliverd-screen": {
"command": "npx",
"args": [
"-y",
"@deliverd/screen"
]
}
}
}Install Deliverd Screen MCP in Codex
$ codex mcp add deliverd-screen -- npx -y @deliverd/screenInstall Deliverd Screen MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.deliverd-screen] command = "npx" args = ["-y", "@deliverd/screen"]
Install Deliverd Screen MCP in Gemini CLI
$ gemini mcp add -s user deliverd-screen npx -- -y @deliverd/screenInstall Deliverd Screen MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"deliverd-screen": {
"command": "npx",
"args": [
"-y",
"@deliverd/screen"
]
}
}
}Automated checks
Ran Nov 14, 2026Full report →Passed our automated checks. That is not a security audit: read the code and give the server the narrowest access it needs. How checks work
- Source repository linked and reachabledeliverd-dev/deliverd-sdk is public and reachablegithub.com/deliverd-dev/deliverd-sdk ↗Pass
- Package names this repository as its source@deliverd/screen names deliverd-dev/deliverd-sdk as sourcenpmjs.com/package/@deliverd/screen ↗Pass
- Build provenance (Sigstore attestation)Sigstore attestation for @deliverd/screennpmjs.com/package/@deliverd/screen/v/0.2.0#provenance ↗Pass
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-12github.com/deliverd-dev/deliverd-sdk ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNo popular package one edit away from @deliverd/screenPass
- No critical advisory in the package or its dependencies1 resolved dependencies checked; no critical advisoryosv.dev ↗Pass
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubDeliverd Screen
A pre-send check for every prompt. Find and redact personal data, secrets and your own confidential terms before text goes to an AI model.
- Local. No account, no key, no network call. The text never leaves your process, unless you turn on the optional Microsoft Purview check, which sends it to your own Microsoft tenant and nowhere else.
- Reversible. Values become placeholders such as
[EMAIL_1](the same value gets the same placeholder throughout), andrestore()puts them back in the model's answer. - Bring your own model for names. Patterns can't find people's names, but a model can. Plug in a Hugging Face token classifier, Presidio, or any model with the same output.
- No dependencies. MIT.
npm install @deliverd/screen
import { screen } from "@deliverd/screen";
const checked = await screen("Reply to mary.byrne@example.ie about IBAN IE29AIBK93115212345678");
checked.verdict; // "redact"
checked.text; // "Reply to [EMAIL_1] about IBAN [IBAN_1]"
const answer = await model.generate(checked.text);
checked.restore(answer); // the real values, back in the answer
What it finds
Email addresses, phone numbers, IBANs, payment card numbers (Luhn-checked),
UK National Insurance numbers, Irish PPS numbers (check character), US Social
Security numbers, IP addresses (v4 and v6), dates of birth, API keys and tokens
(AWS, GitHub, Slack, Stripe, Anthropic, OpenAI, Google, JWTs and more), private
keys, passwords in connection strings and .env files, words
that mark special-category data such as health or religion, and your own
terms.
The four answers
By default the model may receive internal data and no personal data. Pass
allow: { personalData: true, classification: "confidential" } for a model you
have approved for more.
| Verdict | What to do |
|---|---|
allow |
Send text, which is the original. |
redact |
Send text, which has placeholders in place of what the model may not receive. |
ask |
A person has to decide (special-category data, by default). allowed is false. |
deny |
Send nothing. Secrets are refused by default. |
Names: plug in a model
import { pipeline } from "@huggingface/transformers";
import { tokenClassifier, presidio } from "@deliverd/screen/adapters";
const ner = await pipeline("token-classification", "Xenova/bert-base-NER");
const checked = await screen(prompt, { detectors: [tokenClassifier((t) => ner(t))] });
// or a Presidio analyzer you run yourself:
// docker run -p 5002:3000 mcr.microsoft.com/presidio-analyzer
await screen(prompt, { detectors: [presidio({ url: "http://localhost:5002" })] });
A detector is a function from the text to { kind, start, end } spans, so any
model works, including Perplexity's PII-Tracer and GLiNER. Check a model's
licence before you ship it. Run it where the text already is: a detector that
sends the text to a hosted service has already sent it.
Vercel AI SDK
import { generateText, wrapLanguageModel } from "ai";
import { screenMiddleware } from "@deliverd/screen/ai-sdk";
const model = wrapLanguageModel({ model: openai("gpt-5"), middleware: screenMiddleware() });
const { text } = await generateText({ model, prompt });
The middleware screens the whole conversation at once, so one value keeps one
placeholder across messages. It restores the answer, including a streamed one where a
placeholder arrives split across chunks. On a deny or an ask it throws
ScreenBlockedError and the model is never called. It works with AI SDK 7 by
default; pass specificationVersion: "v3" for AI SDK 6.
CLI, pre-commit and CI
cat prompt.txt | npx @deliverd/screen --redact # the redacted text
npx @deliverd/screen --json notes.txt # kinds and offsets, never values
npx @deliverd/screen --fail evals/*.jsonl # exit 1 if anything is found
npx @deliverd/screen --terms clients.txt doc.md # your own words, one per line
In GitHub Actions, uses: deliverd-dev/deliverd-screen@v0 with paths: runs
the same check; for pre-commit, the hook id is deliverd-screen. Both are in
deliverd-dev/deliverd-screen.
MCP
claude mcp add deliverd-screen -- npx -y @deliverd/screen mcp
This gives an agent in Claude Code, Cursor or Codex one tool, screen_text.
It returns the verdict and the redacted text, and never the values it
replaced, so they don't enter the agent's context through the tool.
Microsoft Purview
If your organisation writes its data loss prevention rules in Microsoft Purview, the same rules can apply here, for any model and any tool. Pass a Purview check and the stricter answer wins: a Purview block denies, and a warn is an ask.
import { entraAppToken, purviewCheck, screen } from "@deliverd/screen";
const purview = purviewCheck({
token: entraAppToken({ tenantId, clientId, clientSecret }),
userId: entraObjectIdOfThePerson,
app: { name: "Claims assistant", entraAppId: clientId },
});
const checked = await screen(prompt, { purview });
This is the one thing in the package that uses the network, and only when you
pass it: the text goes to Microsoft Graph (processContent), in your own
tenant, and nowhere else. The Entra app needs Content.Process.All, and the
DLP rule must be created with New-DlpComplianceRule (Purview does not apply a
portal-only rule to an Entra-registered app).
For the MCP server, set DELIVERD_PURVIEW_TENANT_ID,
DELIVERD_PURVIEW_CLIENT_ID, DELIVERD_PURVIEW_CLIENT_SECRET and
DELIVERD_PURVIEW_USER_ID (and optionally DELIVERD_PURVIEW_ON_UNAVAILABLE:
allow, ask or deny).
A net, not a certificate
Structured identifiers are found well. Names are found only as well as the model you plug in. Redacted text about a person is still personal data under GDPR. Use this to send fewer identifiers, never as proof that a prompt was clean.
When a person has to decide
A library can only tell you to stop. Deliverd's
Data Gate makes the call. deliverd.screen() in @deliverd/sdk runs these
same detectors on your machine and sends Deliverd only the counts ("two email
addresses, one IBAN"). The verdict comes back from your organisation's policy and from what
its AI register approves each model to receive. On an ask it puts the
question to the right person in Slack, Teams or on their phone, and every check is
recorded. It takes the same detectors.
Licence
MIT. The detectors are shared with Deliverd's SDKs and held to one set of test
vectors, packages/conformance/screening.json.
Deliverd Screen MCP questions
How do I add Deliverd Screen MCP to Claude Code?
Run claude mcp add deliverd-screen -- npx -y @deliverd/screen in your terminal. Add --scope user to make it available in every project.
How do I add Deliverd Screen MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
Does the data leave my machine?
By default, no. The text never leaves your process unless you enable the optional Microsoft Purview check.
Can it redact names?
Patterns alone cannot find names, but you can plug in a model like Hugging Face or Presidio to handle name detection.
How does it handle redacted values in the model's response?
It uses a reversible process where values become placeholders and the restore() function puts them back into the model's answer.
What happens if a secret is detected?
Secrets are refused by default, resulting in a 'deny' verdict.