Positioning Guardrails

Use this file when writing website copy, docs, release notes, talks, or sales material.

Core Message

Gait is an artifact-first Agent Control Plane for production tool execution.

Runtime governance is often observational; Gait is execution-time decision and proof.

It combines:

• execution-boundary policy enforcement (gate)
• verifiable run artifacts (runpack)
• deterministic incident-to-regression workflows (regress)

What To Claim

• Deterministic verification/diff/stub replay for the same artifacts.
• Offline-first core workflows.
• Default-safe evidence model (reference receipts by default).
• Stable artifact schemas and exit codes as integration contracts.
• Vendor-neutral adapter model across agent frameworks.

What Not To Claim

• "Autonomous AI safety solved" style guarantees.
• Prompt-layer filtering as a complete control model.
• Hosted governance dashboard capabilities in OSS core.
• Real-time fleet control plane features that are not shipped in OSS v1.

Product Boundary Language

Prefer:

• "execution boundary"
• "verifiable receipts"
• "deterministic regressions"
• "incident to regression in one path"
• "camera vs gate: monitor plus enforce"

Avoid:

• "single pane of glass"
• "AI governance suite"
• "black-box risk scoring"

OSS vs Enterprise framing:

• OSS includes hardened runtime enforcement, deterministic artifacts, and local operability gates.
• Enterprise adds fleet-wide policy distribution, org workflows, and centralized governance controls.

Adjacent Stack Language

• "Gait integrates with your existing identity, vault, gateway, and SIEM stack."
• "Guardrails scan content. Gait evaluates structured action intent and enforces policy before execution."
• "Gait produces the evidence your monitoring stack consumes."

Adapter Neutrality Language

• Frame integrations as the same contract across frameworks.
• State explicitly that adapters do not bypass Gate.
• Avoid framework-specific semantics in messaging.