We actively support the latest version of SFSpeckit. Security updates for older versions are not provided; please upgrade to the latest stable release to ensure you have the current security protections.
| Version | Supported |
|---|---|
| v1.1.x | ✅ Yes |
| < v1.1.0 | ❌ No |
Please do not use GitHub Issues to report security vulnerabilities.
To report a vulnerability, please use the GitHub Private Security Advisory feature:
- Navigate to the "Security" tab of the repository.
- Select "Advisories" from the left sidebar.
- Click "New draft security advisory" or use the "Report a vulnerability" button if available.
This ensures that vulnerabilities are disclosed responsibly and allows the maintainers to prepare a fix before the details are made public.
SFSpeckit is designed to be secure-by-default:
- Authentication: SFSpeckit does not store Salesforce credentials. It relies entirely on your local
sfCLI (Salesforce CLI) configuration. - External Requests: The toolkit does not make unauthorized external HTTP requests. All communication with Salesforce orgs is routed through the standard
sfCLI commands. - Data Privacy: No metadata or business data is transmitted to external servers beyond what is required for the LLM to process specific instructions (if you are using a cloud-based AI provider).
All code generated by SFSpeckit follows Article IV: Security-by-Default, which mandates:
- Mandatory use of
with sharingorinherited sharing(neverwithout sharing). - Mandatory use of
WITH USER_MODEfor DML and SOQL. - No hardcoded IDs or secrets.