Operations agents
Coordinate specialized agents across infrastructure, security, incidents, monitoring, and service delivery with bounded tools and budgets.
Inspect agents
Self-hosted governed AI operations
RightAPI Forge
Operate infrastructure, resolve incidents, automate runbooks, and deliver internal tools through one approval-controlled platform.
Community source under AGPL-3.0-or-later. Commercial licensing available.
The running product
Operational work stays visible.
These views come from the deployed RightAPI Forge console. Operators can review live runbook gates, inspect generated tools, and trace release state without leaving the control plane.
Read product documentation
One operational control plane
AI that can act, with controls operators can trust.
RightAPI Forge brings agent coordination, infrastructure operations, incident management, workflow automation, and internal-tool generation into the same governed runtime.
Every consequential action moves through explicit identity, policy, approval, execution, evidence, and rollback boundaries. Models assist the work; the platform owns the controls.
Incident command
Detect, triage, escalate, remediate, verify, and document operational incidents in one attributable timeline.
View incidentsRunbooks and workflows
Turn operational knowledge into repeatable execution with approvals, conditions, recovery paths, and persisted run history.
Open runbooksInternal-tool builder
Generate full-stack operational tools from typed specifications, preview them in isolation, and promote only after quality and security gates pass.
Build a toolGoverned execution lifecycle
From intent to verified outcome.
Autonomy is useful only when execution stays observable and recoverable. The platform records who requested a change, which policy allowed it, what ran, and how the result was verified.
- 01UnderstandTyped intent and operational context
- 02PlanTools, targets, risk, and rollback
- 03ApprovePolicy and human decision gates
- 04ExecuteBounded action with live evidence
- 05VerifyHealth, audit, and reversible release
System architecture
A platform boundary around every model and tool.
Providers are replaceable. Governance, identity, operational state, and evidence remain inside the system you control.
Inputs
Operators
Alerts and events
Service requests
APIs and integrations
RightAPI Forge control plane
Identity and RBACPolicy and approvals
Agent orchestrationRunbooks and workflows
Tool sandboxAudit and evidence
Controlled systems
Servers and clusters
Cloud and SaaS
Data and credentials
Generated tools
Provider-flexible model routing
OpenAI-compatible endpoints
Primary and fallback health control
Model alignment checks
Governance by construction
Controls live below the generated code.
Tooling and agent outputs inherit the platform's security boundary instead of implementing their own ad hoc controls.
Read the security model- Authentication and RBAC
- Tenant-aware identity and permission enforcement for operators, agents, APIs, and generated tools.
- Credential isolation
- Secrets stay in controlled stores and are injected only at approved execution boundaries.
- Approval-controlled change
- Risk-aware gates protect destructive actions, deployments, and sensitive targets.
- Attributable audit
- Requests, decisions, tool calls, releases, and rollbacks produce durable evidence.
v1.0.0-rc.3Public release candidate
1,353automated tests
0known dependency vulnerabilities
Signedevidence and container attestations
Inspect release
Deploy on infrastructure you control
Start with source or a signed container.
Commercial use