Runtime Integrations#

Intended is the authority plane for agent runtimes — it does not replace the runtime. It evaluates a proposed action against policy, returns a decision, issues a signed Authority Token on approval, and records the decision in the audit chain. Intended decides and proves; your agent acts. Intended never holds your keys or calls your systems. This page helps you pick the integration path that fits how much of the runtime you control.

Three patterns#

1. Submit intents directly#

Use this when your code runs before the action and can call the API.

• Submit the action to POST /intent.
• Read the decision: APPROVED (200), ESCALATED (202), or DENIED (403).
• On APPROVED, your agent performs the action itself and presents the authorityDecisionToken to the system it calls.
• Optionally report what happened (POST /intents/:id/outcome); the outcome is appended to the decision's record.

This is the strongest path when you own the integration code. See the API Quickstart and Agent Intent API.

2. Verify the token at your service boundary#

Use this when the system being acted on is yours and you want it to refuse any action that lacks proof of authorization.

• Intended evaluates the requested action and issues a token.
• Your service (or a gateway you run in front of it) checks that token with the verify library — signature, tenant, target, expiry, single-use nonce — and acts only if every check passes.
• Validation is fail-closed; a bad or replayed token is rejected.

See Verify Tokens & Decisions. This is also the shape the NVIDIA OpenShell / NemoClaw policy-compilation flow targets.

3. Intercept tool calls at the gateway#

Use this when actions arrive as model tool calls and you don't want to touch agent code.

• Point the agent's model base URL at the LLM gateway.
• The gateway evaluates each tool call once and applies the decision (forward / deny-rewrite / escalate-rewrite).
• Runs in observe first, then enforce.

Current runtime surfaces#

What Intended does — and does not — do#

Intended classifies and evaluates the action, applies your policy and risk thresholds, issues scoped Authority Tokens, supports escalation and approvals, and preserves audit lineage. It does not take ownership of a third-party runtime, modify its licensing or terms, guarantee the security posture of upstream alpha software, or remove your responsibility to harden the runtime, its credentials, and its allowed endpoints.

Choose your next step#

• LLM Gateway — govern tool calls with one env var.
• Verify Tokens & Decisions — verify the token before your service acts.
• NVIDIA OpenShell / NemoClaw — compile runtime policy YAML.
• Third-party runtime boundaries — security model for upstream runtimes. </content>