Triage open findings
Triage is the routine you run weekly (and on day one of a new tenant). The goal is not "close everything", it is "close the things that matter and decide consciously about the rest".
Step 1, scope
Open the Findings page. Filter:
- Status =
OPEN. - Severity =
CRITICAL,HIGH. Skip MEDIUM and LOW for the first pass. - Project = your current project. (You should already be on it; if not, switch.)
This is your priority list. Typical fresh tenants have 30 to 100 entries here.
Step 2, sort by toxic-combo participation
Click Group by, Toxic combination membership. Findings that sit on at least one chain bubble to the top. These are your highest-leverage fixes.
For each row in the On a toxic combination group:
- Click to open the detail panel.
- Open the Toxic combinations tab in the panel.
- Read the chain. Decide the cheapest break-point (often the public exposure or the leaked secret).
- Apply the remediation. The finding moves to
RESOLVEDautomatically on the next scan.
Closing one finding here often closes 3 to 5 OPEN findings transitively, because the chain disappears.
Step 3, triage the rest by category
Switch to the CSPM module page. Filter Status: OPEN, Severity: CRITICAL, HIGH.
Group by Category.
For each category in turn (IAM, Storage, Network, Database, ...):
- Skim the rules.
- For findings with an obvious fix, apply the fix.
- For findings that are genuine accepted risk, mark
IGNOREDwith a reason. - For everything else, leave OPEN. The next triage session will pick them up.
Repeat for KSPM, DSPM, HCR, IaC, Secrets, Vulnerabilities. Most teams budget 5 to 10 minutes per detector.
Step 4, verify
Go to the Dashboard. Verify:
- The Critical and High counts dropped.
- The Findings trend line for the day is downward.
- Toxic Combinations count dropped (each closed chain).
If a finding moves to RESOLVED and re-opens shortly after, the fix did not take. Re-read the remediation, try again.
Step 5, schedule the next pass
Triage is recurring. A weekly cadence is typical for active tenants. Add a calendar reminder. The goal of each session is to keep the OPEN-CRITICAL count low and the trend flat or downward.
Tips
- Avoid mass IGNORE. It feels productive and is technical debt. Each ignore should have a reason and (ideally) a TTL.
- Track recurring re-opens. A finding that re-opens every week is a process problem, not a technical one. Investigate the deployment pipeline that keeps re-creating it.
- Pair with engineers. When triaging KSPM or IaC findings, work with the team that owns the manifest. Knowledge transfer per session beats one-shot fixes.
What's next
- Resolve a toxic combination, the deep dive
- Suppress or ignore a finding