All capabilities
Capability / 04

Agent runtimes

Tool-use, planning.

Agent runtimes that don't just call APIs but reason about what they returned. Tool-use, planning loops, verification, and the structured failure modes you need before an agent touches production.

§ 01

Where agents are real, and where they aren't yet

Agents are real in narrow, well-instrumented domains with clear success criteria and reversible actions. They are not yet real for open-ended, irreversible decisions in unfamiliar systems. We build for the first category and we're honest about the second — including with our own portfolio companies.

§ 02

What we build into agent runtimes

01

Tool registry

Typed, versioned tools with explicit auth, rate-limit, and side-effect contracts.

02

Planner

Decomposition strategies suited to the domain — tree, graph, or replanning loop.

03

Verifier

A second pass that checks the work before the agent commits an action.

04

Memory

Episodic, semantic, and working memory — each with explicit eviction policy.

05

Sandboxing

Capability-scoped execution. The agent can only touch what the operator allowed.

06

Human-in-the-loop

Confidence-scored escalation. The agent knows when to stop and ask.

§ 03

What we ship

Bounded
Action surface — by design
Traced
Every step, every tool call
Reversible
Defaults wherever possible

Ready to look at this
in your context?

Start a conversation