Workspace Chat
Not a chatbot. The interface to the platform.
Workspace Chat queries your actual findings, scans and knowledge graph. It is a way to interrogate the platform in plain language, not a general assistant bolted onto a dashboard.
The distinction matters
A general assistant can describe what SQL injection is. That is not the question an engineer has at eleven at night. The question is whether this finding, in this service, on this code path, is reachable and worth stopping the release for.
Answering that requires the finding, the scan that produced it, the graph relationships behind the rule, and the deployment context around the service.
Workspace Chat sits on top of all of it. Answers cite the findings and evidence they came from, so you can check the reasoning instead of trusting it.
What you can ask it to do
Explain finding
What matched, the path it followed, and what it means for this service specifically.
Generate Jira
A ticket carrying the reasoning, evidence and suggested fix, not just a title and a severity.
Suggest fix
A concrete change for the affected code, in the language and framework it is written in.
Investigate exploitability
Whether the vulnerable path is reachable and exposed in this deployment, and what would have to be true to exploit it.
Explain CVE
Affected ranges, exploitation conditions, and the upgrade path that resolves it here.
Explain rule
What the detection tests for, and the conditions under which it legitimately does not apply.
Generate evidence
Control evidence assembled from scans that already ran, traced to their source.
Security Q&A
Questions about your posture answered from your data rather than from general knowledge.
Cross-finding reasoning
Whether separate findings combine into an attack path, or whether one fix closes several at once.
Questions it answers from your data
Each of these requires joining findings to context: the kind of question that otherwise becomes an afternoon of manual correlation.
- Which of our critical findings are actually reachable from the internet?
- Is this CVE exploitable given how we use the package?
- What would one upgrade of this dependency close?
- Show me every finding that maps to OWASP A01 in the payments service.
- Why was this severity downgraded, and by whom?
Answers carry citations
Every response links the findings, scans and graph nodes it drew on, so the reasoning can be verified rather than taken on faith.
Decisions stay human
Chat assembles context and proposes options. Accepting a risk or downgrading a severity remains an action a person takes and signs.
Scoped to your workspace
It reads the findings you have access to and nothing else: the same permissions that govern the console govern chat.