The job of this console
Read the hint, open the log, write or dismiss. Do not leave the only copy in chat.
AI · Risk
Risk Agent
Suggested risks and concentrations — you still write the row on Projects → Risks (or issues/RCA).
Scroll to Flowcharts. Scenario: Take a risk suggestion to the project log.
Risk hints
A suggestion is not a logged risk
The risk agent points at concentrations. The project risk (or issue) row is still the record.
AI Intelligence → Risk agent
Route
/ai/risk-agent
Org flag riskEnabled
Project Manager · Executive
Seat at the table
Project managers and executives with riskEnabled. The agent does not bypass project access.
The working rhythm
Scan, file or skip, continue delivery work.
Moments that bring you here
Before steering. After a slip. A customer severity. When analytics heat looks new.
ⓘ
Log remains source
Masked audit only. NPD gates are a different process.
Flowcharts
1 — Suggest then file
Read a hint, write the risk on the project.
flowchart TD
A[Open Risk agent] --> B[Suggestions]
B --> C[Project risks]
C --> D[Write / update row]
D --> E[Analytics pack later]
- 1OpenAI Intelligence → Risk agent.
- 2ReviewConcentrations.
- 3FileProjects → Risks.
- 4Or discardIf the source disagrees.
Overview
Risk agent highlights possible risks from project and commercial signals. The risk record stays on the project. Analytics risk intelligence is the pack. Work Graph nodes stay masked.
How to use
1. Do not file duplicates
Search the log first.
2. Issues vs risks
Use the right register.
3. Mask customer secrets
In prompts and snippets.