An asset inventory created from what is actually running and connecting in production, rather than what was declared at design time. For AI agents, it captures processes, model endpoints, API use, and network paths, which makes governance evidence-based instead of aspirational.
Expanded Definition
Runtime-derived inventory is the discipline of building an authoritative asset picture from production telemetry, not from CMDB entries, deployment manifests, or architecture diagrams. In NHI and agentic AI environments, that means discovering what is actually active: service accounts, workloads, model endpoints, API calls, tool connectors, and the network paths those entities use.
This matters because inventory quality determines whether governance is evidence-based or merely assumed. A runtime-derived view often complements a traditional inventory, but it is not the same thing. Definitions vary across vendors on whether the inventory should include only live processes or also dormant-but-deployed components, so teams should document scope explicitly. The most useful reference model is the one that ties discovery to operational behavior, as reflected in the NIST Cybersecurity Framework 2.0 approach to continuous asset understanding.
The most common misapplication is treating a deployment list as a runtime inventory, which occurs when teams assume anything approved for release is still present, reachable, and in use.
Examples and Use Cases
Implementing runtime-derived inventory rigorously often introduces telemetry and collection overhead, requiring organisations to weigh operational visibility against added complexity in production systems.
- A platform team discovers that an AI agent is still calling a deprecated model endpoint after the documented migration window, so the runtime inventory becomes the trigger for decommissioning and access cleanup.
- A security team compares observed service-account activity against declared ownership and finds an orphaned workload still authenticating from a forgotten container cluster, which helps reduce hidden NHI exposure.
- During an access review, analysts use runtime evidence to confirm which secrets, tokens, and certificates are actually exercised, rather than relying on stale spreadsheets or ticket history. This aligns with the visibility and lifecycle concerns described in the Ultimate Guide to NHIs.
- An AI governance team maps active tool usage to approved policy boundaries and identifies a model workflow reaching out to an unapproved API, then blocks the path before broader sprawl develops.
- An incident responder reconstructs the live dependency chain for a compromised agent and uses observed network paths to determine which adjacent systems may have been touched, rather than inferring from design documents alone.
For organisations building this capability, the practical benchmark is whether the inventory can answer “what is active right now?” using observed evidence, a principle that also supports continuous control validation described in the NIST Cybersecurity Framework 2.0.
Why It Matters in NHI Security
Runtime-derived inventory closes a major blind spot in NHI security because non-human identities change faster than most governance records do. That gap is especially dangerous when service accounts, API keys, and agent tool paths persist after application changes, since stale records can hide excessive privilege, orphaned access, and unreviewed dependencies. NHI Mgmt Group notes that only 5.7% of organisations have full visibility into their service accounts, which shows how often governance starts from incomplete evidence rather than operational reality. The same visibility issue is central to the findings in the Ultimate Guide to NHIs.
In practice, runtime-derived inventory supports better rotation decisions, faster offboarding, tighter least privilege, and more credible Zero Trust segmentation. It also helps teams detect when an AI agent begins to use new endpoints or unexpected integrations, which is often the first sign of configuration drift or compromise. Without it, organisations tend to discover that an identity or tool chain existed only after a breach, an outage, or a failed audit, at which point runtime-derived inventory becomes operationally unavoidable to reconstruct what was actually connected.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
OWASP Non-Human Identity Top 10, CSA MAESTRO and OWASP Agentic AI Top 10 address the attack and risk surface, while NIST CSF 2.0 and NIST Zero Trust (SP 800-207) set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| OWASP Non-Human Identity Top 10 | NHI-01 | Runtime inventory supports discovery and visibility of all non-human identities and their activity. |
| NIST CSF 2.0 | ID.AM | Asset Management requires understanding what is present and operating in the environment. |
| NIST Zero Trust (SP 800-207) | Zero Trust depends on knowing current subjects, devices, and connections before granting access. | |
| CSA MAESTRO | Agentic security frameworks emphasize observing live agent behavior, tools, and dependencies. | |
| OWASP Agentic AI Top 10 | Agentic AI guidance highlights runtime monitoring to detect unexpected tool use and drift. |
Continuously enumerate live NHIs and reconcile them against approved records to remove blind spots.
Related resources from NHI Mgmt Group
Deepen Your Knowledge
Reviewed and updated by the NHIMG editorial team on August 19, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org
