Sanction

Demo — Sanction HQdemo

OverviewApprovalsSeatsPoolsPolicySpendAuditProvidersCredentialsTeam & access
Signed ledger
demoLog in
Sanction
demoLog in
OverviewApprovalsSeatsPoolsPolicySpendAuditProvidersCredentialsTeam & access

Policy

One decision engine governs spend, tools, and capability. Everything here is checked before an agent acts, and every change is a replayable revision.

Tool posture

0 blocked · 0 escalated

Capability rules

0 active rules

Token cost month

$0.00

Approval pressure

0 pending · 0 escalated

Policy packs

Curated baselines along the governance ladder. Preview replays a pack over your last 30 days before you commit; applying replaces the current policy.

Metering

Watch first — everything passes, everything is measured.

Metering first

Watch before you govern — everything passes, everything is measured.

Your first week with agents: see real spend patterns before choosing limits.

Gateway token budget

Start with model-call metering and a hard stop before the provider sees over-budget traffic.

Teams already using LiteLLM, Vercel AI SDK, OpenAI-compatible clients, or a homegrown LLM gateway.

Authorization

Ceilings — hard limits without a committee.

Startup defaults

Sane guardrails for a small team's first production agents.

A few agents doing real work; you want a ceiling, not a committee.

MCP tool governance

Put allow, escalate, and deny rails around the MCP tools your agents discover.

Teams adding Sanction to Cursor, Claude Desktop, Claude Code, Windsurf, or custom MCP hosts.

Governance

Humans in the loop where it matters.

Team workspace

Departmental budgets, human escalation where it matters.

A team's shared wallet: real budgets, approvals for the unusual, room to work.

Coding agent seat

Let coding agents read and propose freely; escalate writes, deploys, and new tools.

Cursor, Claude Code, Codex, Cline, and OpenHands agents moving from sandbox to real repos.

Fleet channel envelope

The pool policy for one spending channel of an agent fleet — monthly envelope, pooled token cap, escalation line, outcome-ceiling ready.

Ops running a fleet where channels (paid media, content, outbound) are delegated pools and every seat spends. Apply to the channel pool, not the root; set outcome_kind + a ceiling once you report outcomes.

Evidence

Fail closed, prove everything.

Compliance baseline

Fail closed — everything above a whisper asks a human.

Regulated work: every spend escalates, every new capability escalates, timeouts deny.

Client-safe launch

A launch policy an AI agency can hand to a client: visible spend, approval for risky work, evidence after.

AI agencies and consultants shipping agents into client accounts this week.

Payment agent mandate

Hold policy and evidence in front of whichever payment rail the agent uses.

AP2, x402, checkout, procurement, and purchasing-agent pilots where money movement needs a mandate.

No egress

Only on-box tools pass — every cloud call is denied and logged for the assessor.

Sanction Local installs: regulated practices running local models where data must not leave the building.

Policy

Budgets

Hard spending lines, checked before every action.

Human approval

The escalation window: silent below one line, paused for a person above the other.

Categories & tools

What agents may buy and which tools they may call — blocked always wins.

Capability rules (0/200)

Governs what agents may acquire — installing skills, adding plugins, reaching new APIs. Patterns prefix-match with *; block wins, then the allow list, then escalate.

No capability rules — acquiring skills/plugins/APIs is ungoverned. Add a rule like skill:install:* → escalate.

Viewing the demo — log in to edit your own policy.Log in