Application Security Engineer Working With SIRT
This runbook is meant to help AppSec engineers who need to engage and work with SIRT to respond to a HackerOne report, discovered vulnerability, or other security incident.
Requirements for using /security to engage SIRT
If this is a P1S1, follow the P1S1 runbook and engage the security on-call.
When using /security
to engage SIRT:
- Only use
/security
after we have validated the vulnerability or otherwise confirmed the incident - Fill out the form as completely as you can
- Include a link to the issue or HackerOne report, a summary of what has occurred, and a description of what we are asking SIRT to help with
- Select the appropriate urgency (see below for some guidelines)
Situations where SIRT needs to be engaged
Note: This is not an exhaustive list, there may be other situations in which you need to engage SIRT.
- P1 vulnerabilities
- Publicly known P2 vulnerabilities with a publicly available exploit
- Secret leaks
- Vulnerabilities that could result in account takeover
- Vulnerabilities that lead to customer communications being required in addition the usual processes
- Security fixes made in public that should have been made against a security repository
- Personal data leaks
- DNS takeovers
Urgency guidelines
These are some guidelines for selecting the urgency in the /security
form:
Type of incident | Urgency | Notes |
---|---|---|
Any kind of S1 | Urgent (page right away) | Examples include critical vulnerabilities, exposure of red data, still-valid team member token leaks |
Personal data leaks | Not Urgent (review within 24 hours) | Could be Urgent depending on the volume and the data |
Public merge requests fixing vulnerabilities not labeled security-fix-in-public |
See notes | S2 and above is likely Urgent, S3 and below is Not Urgent |
For more information, consider viewing the SIRT incident classification and severity matrix page.
Getting SIRT attention without using /security
In some situations, we just want SIRT to be aware of something that is happening or may become an incident soon. An example of this would be a high severity unverified HackerOne report from a reliable HackerOne reporter.
Put a message in the #sd_security_operations
Slack channel with a brief description of the situation. You may consider using /sirtoncall
to determine who is on-call and pinging them.
For HackerOne reports, you can also import the report into a GitLab issue and then mention @gitlab-com/gl-security/security-operations/sirt
on it. Keep in mind this may not be appropriate, since it will ping all team members of SIRT.
Assisting with incidents in general
- Be ready to join the SIRT bridge or collaborate with SIRT asynchronously
- Assist in writing summaries, documenting reproduction steps, and keeping the timeline up to date
- Add information and context in the incident issue and discussion threads
- Update the incident issue with access timestamps, your IP address, and the IP address of any HackerOne reporters (and/or HackerOne triage team) to assist with log reviews
46417d02
)