EU AI Act evidence, generated
Article 13 reports auto-mapped from every agent decision. One-click PDF export. Always audit-ready.
POST /v1/compliance/reports/eu-ai-act
{ "from": "2026-01-01", "to": "2026-07-31",
"format": "pdf" }EU AI Act · August 2026 enforcementGovern every AI agent
Govern every AI agent
you ship.
RuneSignal is the governance, accountability, and operational control plane for AI agent fleets in production — compliance evidence, human approvals, and full inventory from day one.
SOC 2 Type II in progressEU-hosted deploymentEd25519-signed audit trailNo credit card required
142+
Agents governed
Per tenant, median
17
Shadow agents found
Per deployment rollout
< 40ms
Firewall latency
p95 decision time
5+
Frameworks mapped
Out of the box
Five primitives. One governance plane.
Every capability is an API, not a slide deck. Ship governance the same way you ship code — with full observability and zero ceremony.
Human-in-the-loop approvals
REST API with automated blast-radius scoring on every pending action. OpenAPI spec included.
POST /v1/exceptions/{id}/resolve
{ "decision": "approved",
"blast_radius_score": 7.4 }Inventory + shadow AI discovery
Every declared agent, plus the undeclared ones. Risk scoring, ownership, compliance — all in one place.
GET /v1/agent-inventory?include_shadow=true # 142 agents · 17 shadow · 3 high-risk
Slack, ServiceNow, Jira
Unified adapters route HITL approvals, incidents, and tickets from a single API call.
POST /v1/integrations/slack/notify
{ "channel": "#ai-ops",
"exception_id": "exc_9f2a4b" }SDK + LangChain plugin
Instrument any agent in three lines. Full TypeScript types, zero-config tracing, automatic policy enforcement.
import { RuneSignalClient } from '@runesignal/sdk';
rs.instrument(myAgent);Four steps to audit-ready.
From first agent to signed compliance report — in an afternoon.
- Step 01Register agentsDeclare every AI agent — or let Shadow Discovery find the ones you missed.
- Step 02Define policiesBlast-radius thresholds, frameworks, HITL rules, and data residency requirements.
- Step 03Route through the firewallEvery agent action passes real-time allow / block / escalate decisions.
- Step 04Ship evidenceOne-click compliance reports and signed audit exports — always current.
Built for the frameworks that matter.
RuneSignal maps agent behavior to six compliance frameworks out of the box. No custom configuration required.
EU AI Act
Article 13 · August 2026
SOC 2 Type II
In progress
GDPR
EU data residency
ISO 27001
Controls mapped
NIST AI RMF
Full framework
OWASP LLM
Top 10 mapped
Works with the stack you already run.
Native adapters for every major AI framework and enterprise ticketing system.
LangChain
OpenAI
Anthropic
AutoGen
CrewAI
Slack
ServiceNow
Jira
GitHub
PagerDuty
Depth where it counts.
Most AI governance tools guard the prompt. RuneSignal governs the agent.
| Capability | RuneSignal | Credo AI | Lakera |
|---|---|---|---|
| EU AI Act Article 13 evidence | Yes | Partial | No |
| HITL API with blast-radius scoring | Yes | No | No |
| Shadow AI discovery | Yes | No | No |
| Slack / ServiceNow / Jira adapters | Yes | Partial | No |
| Developer SDK + LangChain plugin | Yes | No | Partial |
| Real-time action firewall | Yes | No | Yes |
| Cryptographic audit trail | Yes | No | No |
| Multi-framework compliance (5+) | Yes | Yes | No |
Ship governance
in an afternoon.
Free to start. EU-hosted. Signed audit trail from day one. No credit card required.