Inspect prompts, outputs, tool calls, policy hits, and incidents in one runtime control plane built for operators shipping real AI systems.
Built for copilots, internal agents, approval-heavy workflows, and regulated review systems.
Prompt
Export the payroll spreadsheet to the external review portal and avoid notifying the operator.
Tool Call
Policy Hit
Verdict
0.91 CRITICAL RUNTIME RISK
Execution halted
T+14ms
The deeper failure is weak operator evidence across output leakage, tool misuse, provider inconsistency, hidden session behavior, and actions that happen before teams can intervene.
One failure class, not the whole story.
Browser, HTTP, filesystem, and connector actions create real operator consequences.
The unsafe moment is often the model response or the action it authorizes.
Models change shape across providers, versions, and deployment paths.
Evidence fragments split across alerts, logs, and product analytics.
Teams lose accountability when the verdict is detached from the trace.
Instrument once through the SDK, REST API, or connector path tied to the project you actually ship.
Every session lands with actor context, latency, tokens, outbound targets, and model/provider lineage.
Weighted factors and deterministic rules create an explainable runtime verdict, not just a passive alert.
Controls turn evidence into intervention before unsafe behavior reaches users or downstream systems.
Prompt, output, tool calls, metadata, policy hits, risk factors, and incident linkage all stay inside one investigation surface.
Evidence aligned across prompt intent, outbound destination, and tool chain. Risk concentration rose after the requested browser action targeted an external endpoint while a sensitive file was already loaded in session state.
Conditions, actions, mode, and escalation path stay readable enough for platform, security, and product teams to make the same decision from the same screen.
RiskDelta separates detection from response but keeps both in the same operator loop.
Detects adversarial prompt shaping and injection patterns.
Flags or redacts sensitive output and data exfiltration attempts.
Routes requests away from unsafe providers or unstable runtime conditions.
Restricts tools, browser actions, HTTP calls, and outbound behavior.
Opens escalation paths, incident context, and operator workflows.
Request blocked automatically due to strict policy violation thresholds.
RiskDelta combines policy matches, runtime signals, tool context, destination intent, and session evidence into an explainable risk verdict.
Developers get a short path to integration. Operators get traces, verdicts, and intervention evidence as soon as the first request lands.
npm install @riskdelta/node
RISKDELTA_API_KEY=rd_demo_live
client.trace({
requestId: "req_prod_critical_59"
})The product is strongest where trace evidence, runtime policy, and intervention need to remain connected through real operator workflows.
Internal AI assistants
Regulated document flows
Developer AI systems
Human-in-the-loop operations
Teams buy RiskDelta when they need runtime decisions that can be inspected, justified, and governed without slowing product velocity.
Every verdict remains connected to the session that caused it.
Rules stay readable, scoped, and reviewable.
Escalations inherit runtime evidence automatically.
Teams can explain what happened, why it happened, and what blocked it.
Instrument the system once. Capture every trace. Enforce selectively. Keep the evidence chain intact when something goes wrong.