@flow-scanner/lightning-flow-scanner-core
Version:
A lightweight engine for Flow metadata in Node.js, and browser environments. Assess and enhance Salesforce Flow automations for best practices, security, governor limits, and performance issues.
24 lines (15 loc) • 1.63 kB
Markdown
# Security Policy for Lightning Flow Scanner
## Security Practices
- Code is open-source and peer-reviewed by the community.
- Vulnerabilities can be reported privately via [GitHub vulnerability reporting](https://github.com/Flow-Scanner/lightning-flow-scanner-core/security).
- All changes are scanned with [Snyk](https://github.com/snyk/cli) prior to publication.
- Releases are published to npm using **GitHub Actions Trusted Publishing (OIDC)**.
- Tags (`v*`) trigger automated `npm publish`, providing a full audit trail.
## Data Handling
This tool collects zero user data. No credentials, PII, payment info, health data, or user content is ever stored, transmitted, or shared. All analysis runs 100% client-side with no network calls to external services.
We temporarily use metadata (e.g., Flow metadata, timestamps) in-memory only for real-time functionality during your session. This data is never stored, logged, or transmitted and is discarded immediately when the session ends.
## Dependencies
We actively track and maintain an up-to-date inventory of all third-party dependencies to ensure security and compatibility. Our dependencies include:
| Package | License | Purpose |
| ----------------- | --------------------------------------------------------------------------------- | ---------------------------------------------- |
| `fast-xml-parser` | [MIT](https://github.com/NaturalIntelligence/fast-xml-parser/blob/master/LICENSE) | Validate XML, Parse XML and Build XML rapidly. |