Human-triggered
Add a “Report to Agent911” action inside your product when a user encounters suspicious or harmful behaviour.
Embedded reporting
When something goes wrong inside your app, marketplace, CRM, support flow or agent workflow, let the user—or your system—report it directly to Agent911. Reporter identity can stay private publicly.
Add Agent911 to your workflowReport where the problem happens
A bad customer interaction, suspicious seller, unsafe agent action or fraud event should not disappear inside one company’s system. Agent911 turns it into a structured report that can contribute to wider network intelligence.
Add a “Report to Agent911” action inside your product when a user encounters suspicious or harmful behaviour.
Let fraud, risk, support, moderation or agent systems create a report when your own rules detect a problem.
Pass the relevant transaction, interaction, URL, screenshot, file or event context with the report so the signal is useful from the start.
Developer integration
Use the integration surface that fits your product. Keep reporter identity private publicly while preserving the provenance needed for deduplication and abuse controls.
Create report records from apps, marketplaces, CRMs, support tools and server workflows.
curl -X POST https://www.agent911.ai/api/v1/reports \
-H "Authorization: Bearer a911_live_…" \
-d '{"target":"@account","category":"impersonation"}'Give compatible assistants and agents a structured Agent911 reporting tool.
{
"mcpServers": {
"agent911": { "url": "https://www.agent911.ai/mcp" }
}
}Report from terminals, developer workflows and CI pipelines.
npx @agent911/report @account \ --category impersonation \ --evidence ./screenshot.png
Agent911 captures reports. Clawdbase remains the destination for current Trust Score + Confidence Score.