Skip to main content
Findings arrive from scanners, pentests, bug bounty programs, and advisories faster than anyone can review them. Each source has its own queue and its own severity scale, and the same issue often arrives from several of them under different titles. Someone still has to read every report, work out whether it reproduces, and reply to the reporter, which is why most backlogs grow faster than they shrink. A source’s severity score doesn’t help much either. It describes how bad an issue could be, not whether it is reachable in your environment.

How Neo Solves This

Triage pulls every source into one backlog and works it down to zero. Neo tests each finding against your environment and gives it a verdict: confirmed with proof, merged into an existing finding, or closed with the reasoning attached.
  1. Pulls every source into one backlog. Bug bounty platforms, GitHub, and security mailboxes connect natively. Scanners and any other tool with an API connect from their API docs. Neo only reads from sources and never changes anything in them. Every finding lands in one backlog, where the board moves it through Untriaged, Investigating, Verdict ready, In fix, and Closed. Switch to the list view and use filters to narrow it down, or open a finding to see the original report as the source sent it.
  2. Deduplicates across sources. Deduplication runs across all sources. The same issue reported by three tools under three different titles becomes one finding, and the copies are filed as duplicates.
  3. Reproduces instead of reading. Neo does not judge a report by how convincing it reads. It works out what it would take to reproduce the issue in your environment, then tries. When Neo confirms a finding against a running target, it is because it saw the issue happen. Code-only findings are traced through the code and validated live when a target is available. See Verification for how Neo proves findings.
  4. Needs no prompt. Click Triage on a finding, or drag it into Investigating on the board, and Neo starts a task with a purpose-built triage workflow. Open the task to watch Neo read the report’s claims, check for duplicates, and test the issue live. To triage many findings at once, multi-select them or select all, then triage.
  5. Leaves the decision with you. When triage finishes, the finding moves to Verdict ready with Neo’s full triage report and a verdict. Neo’s severity reflects what is reachable in your environment, and the verdict shows it beside the source’s score. You accept, merge, or decline, and you can challenge any verdict. See Review verdicts.
  6. Keeps the backlog at zero. Turn on auto-triage for a source to have Neo check for new findings on a schedule, from every 15 minutes to once a day, and triage them as they arrive. The rule runs as the person who saves it. Insights tracks the backlog over time: what is open, what is past its SLA target or due soon, the backlog by age and by source, the median time to close, and how long until the backlog reaches zero at the current pace.

What This Looks Like in Practice

It’s Monday morning and 40 new findings arrived over the weekend: HackerOne reports, a Snyk sync, and a handful of emails to your security mailbox. You open the backlog, select all of them, and click Triage.
The Triage board with findings moving from Untriaged to Investigating and Verdict ready
Neo files 11 as duplicates, including one SSRF that a researcher, Snyk, and a mailbox report each described under a different title. It closes 18 as Not applicable or Informative with the reasoning attached, marks 2 as Spam, asks 3 reporters for more information, and confirms 6 as Valid. One of the six is a HackerOne report the researcher rated Critical. Neo reproduced it, but the endpoint only answers authenticated admins, so the verdict shows Neo’s Medium beside the source’s Critical. Before you accept the lower severity, you open the finding’s Activity and ask:
Neo follows up with the proof. You accept the six valid findings into In fix, send the replies Neo drafted for each reporter, and decline the rest. Then you turn on auto-triage for each source, so next weekend’s findings have verdicts waiting on Monday.

What You Get

  • One backlog for every source. Bug bounty reports, advisories, mailbox reports, and scanner findings in one place, with duplicates merged across sources.
  • Verdicts backed by reproduction. A valid finding comes with evidence from your environment, not a restatement of the report.
  • Severity that reflects your environment. Neo’s severity sits beside the source’s score, so you can see where they differ and why.
  • Replies drafted for you. Every valid finding comes with a reply to the reporter ready to send.
  • A backlog that stays at zero. Auto-triage handles new findings as they arrive, and Insights tracks SLA targets and time to zero.

Connect a source

A source is any system that produces findings: a bug bounty platform, GitHub advisories, a security mailbox, or a scanner. Add one from Add a source in Triage. Bug bounty platforms, GitHub, and security mailboxes connect natively. Scanners and other tools connect from their API docs. Neo only reads from sources and never changes anything in them.

HackerOne

Create an organization API token under Organization Settings > API Tokens in HackerOne (docs), enter the token identifier and token in Triage, and pick a program. Its reports sync into the backlog.

Bugcrowd

Create an API credential on the Bugcrowd API credentials page, enter the identifier and secret in Triage, and choose the programs to sync.

GitHub Security Advisories

Install and claim the Neo GitHub App (see Install & Connect), then pick a repository. Its security advisories sync into the backlog.

Security Inbox

Sign in with Google so Neo can read the reports sent to your security mailbox. In the mailbox settings, add a triage policy describing what makes a report valid (scope, out-of-scope issues, minimum impact), an optional inbox filter as a Gmail search such as to:security@acme.com, and choose whether the whole team or only you can see it.

Scanners and other tools

Scanners and any other tool with an API connect from their API docs. That includes issue trackers like Jira or Linear when the findings you want to triage already live there as tickets. You describe the API, and Neo reads it, finds the endpoint that returns findings, and builds the integration itself. For example, to bring in issues from Snyk:
1

Save the API token as a secret

Create a Snyk API token and add it under Settings → Secrets. If you paste a key into Triage instead, Neo offers to save it as a secret and keeps it out of the source configuration.
2

Describe the source

In Triage, click Add a source. Paste a link to Snyk’s API documentation, a curl command that lists issues, or a plain description of what to import, and attach a spec or README if you have one. Pick the Snyk secret, then click Connect.
3

Follow setup

The source pane shows Neo’s progress as it reads the docs, chooses the endpoint that lists issues, probes it with a sample request, and maps the response to findings. If Neo needs something from you, such as which organization to import, it asks and gives you options to pick from. If setup fails, try again or edit the configuration by hand.
4

Check the mapping

Open the source’s settings to review the sign-in type, the request, paging, and the field mapping for title, severity, and other fields. Fix anything Neo got wrong.
5

Import and keep it synced

The first sync imports existing issues. Large backlogs take longer the first time. Set Sync frequency, from every hour to once a day, so new issues keep arriving, or click Sync now at any time.
6

Triage and automate

Triage a few findings by hand to check the verdicts, then turn on auto-triage for the source.

Review verdicts

A finding in Verdict ready carries one of these verdicts.
  • Valid: reproduced in your environment, with evidence, a severity based on what is actually exposed, and a reply to the reporter drafted for you to send
  • Likely duplicate: matches an existing finding from any source
  • Needs more information: Neo could not reach a verdict without more detail from the reporter
  • Informative: accurate, but not an exploitable risk in your environment
  • Not applicable: the claim did not hold up, with the reasoning attached
  • Spam: not a genuine report
A finding in Verdict ready is waiting on your decision in Triage. Accept real issues and track the fix in In fix, merge duplicates, or decline the rest as a false positive, accepted risk, or won’t fix. Neo’s severity reflects what is reachable in your environment, so it often differs from the source’s score. The verdict shows both side by side. If you disagree with a verdict, open the finding’s Activity to follow up with Neo directly. Ask for more proof, challenge the reasoning, or jump back into the underlying task.

Setup

To start triaging:
  1. Open Triage, click Add a source, and connect your first source. Setup for each one is in Connect a source.
  2. Triage a few findings by hand and review the verdicts.
  3. Turn on auto-triage for each source once you trust the verdicts.
  4. Set SLA targets per severity to match your team’s remediation policy.

Get started with Triage

Get your first finding triaged in a few steps.

Verification

How Neo proves a finding before it reports it.