Use case · GitHub Actions
Clean pull request text before an AI step reads it.
An AI review or summary step in GitHub Actions reads pull request descriptions, and those carry customer emails from bug reports and keys pasted to show a failure. A workflow step sends the text through the Vrymdor™ AI Node first.
The problem
An AI step in a pipeline reads whatever text the workflow hands it.
- A pull request description quotes the bug report, with the customer’s email address in it.
- A teammate pastes a token into the description to show which key stopped working.
- A description can carry a phrase written to take over the AI agent that reads it, such as “ignore previous instructions”.
With the AI Node
A step sends the pull request text to your node and writes the cleaned version to a file the next step reads. Your dashboard walks you through it.
- In your dashboard, open Protect a GitHub repository and press Create a GitHub agent.
- In the repository on GitHub, add the three values as repository secrets (Settings, Secrets and variables, Actions): VRYMDOR_NODE_URL, VRYMDOR_AGENT_ID and VRYMDOR_AGENT_SECRET.
- Add the workflow file the dashboard gives you as .github/workflows/vrymdor-node.yml and commit it.
- Open a test pull request. The sanitize check shows [REDACTED:EMAIL] in place of an email address.
The dashboard’s GitHub setup uses the hosted node (Pro and up). On the free plan, the guide shows the same workflow with a node your runners can reach.
Try it on a pull request description
The AI Node hides email addresses, phone numbers, card numbers, US Social Security numbers, private keys, and the API keys and tokens for OpenAI, Anthropic, Stripe, GitHub, AWS, Google and Slack, plus bearer tokens and JWTs. Each one becomes a label like [REDACTED:EMAIL], so the model still knows something was there.
The title and the issue link stay, so the AI step still knows what the change is about. Masked here: EMAIL 1, GITHUB_TOKEN 1, AWS_ACCESS_KEY 1.
This demo uses the Playground's Local engine: public, well known detectors set to the types the AI Node masks by default, in the AI Node's [REDACTED:TYPE] format. In your workflows, the Vrymdor™ AI Node does the masking.
What it does not cover
Plainly, so you know where you still need care.
- Names and street addresses are not masked by default. On Pro and Business you can add your own masking patterns for details with a fixed shape.
- It works by pattern matching on known formats, with the Luhn test for card numbers. A detail written in an unusual way, such as a card number spelled out in words, can get through.
- It hides private details before the model reads the text. It does not check what the model answers, and it does not make an AI model safe in general.
Start free, upgrade when you grow.
Community is free for up to 2 agents and 10,000 requests a month, with a hard cap so nothing is ever billed. Pro is $29 a month for up to 15 agents and 100,000 requests, and Business is $99 a month for up to 50 agents and 500,000 requests, each with 14 days free. Enterprise is sized to your contract.