Home · Platform · Agent Accountability
Platform · Agents

When an AI acts on its own, know exactly what it did.

Under whose authority, at what cost, and be able to pause or roll it back, with evidence. AskLedger gives every production agent a stable machine identity and a named human owner, keeps its capabilities inside pre-approved limits, and records a signed receipt for every action it attempts and every action it executes.

The control model

Six controls, on every production agent.

Autonomy without accountability is a liability. The control model keeps an agent's blast radius small, its authority explicit, and its runtime behavior tied back to the decisions and configurations that permitted it.

01

Identity & owner

Every production agent has a stable machine identity and a named human owner accountable for it.

02

Pre-approved capabilities

Tools, data classes, jurisdictions and spending limits are approved before the agent runs.

03

Policy before action

Policy is evaluated before sensitive actions, with attempted and executed events recorded separately.

04

Least privilege

Short-lived credentials and least privilege keep the blast radius of any single agent small.

05

Tested controls

Escalation, pause, revocation, rollback and kill-switch controls are tested and evidenced, never assumed.

06

Linked runtime

Runtime activity links to evaluation results, versioned configurations and downstream outcomes.

The agent receipt sequence

Every agent action leaves an evidence trail.

From the moment a task is assigned to the moment it's completed, or rolled back, each step emits a receipt or evidence event, so the full story of what the agent did is reconstructable and verifiable.

01

Task assigned

Requester, purpose, agent identity and approved use case.

02

Plan created

Plan hash, tools anticipated and the policy version in force.

03

Tool request

Target resource, requested scope and the policy decision.

04

Action executed

Result, changed resource, cost and the execution identity.

05

Human escalation

Reviewer, decision, and any exception or override.

06

Completion

Outcome, quality signal and business reference.

07

Rollback / revocation

Control invoked, its result and the recovery state.

Protocol alignment

Built on the protocols agents already speak.

AskLedger doesn't invent a parallel agent stack. It captures interaction metadata from the protocols agents use to talk to tools and to each other, and relies on your existing workload identity patterns where possible.

MCP

Supports Model Context Protocol interaction metadata: tool requests, policy decisions and execution results captured as agents call their tools.

Agent-to-agent (A2A)

Captures agent-to-agent interaction metadata, so multi-agent handoffs stay attributable across the chain.

This aligns with emerging guidance that emphasizes agent ownership, inventory, identity, observability, data controls and standard protocols such as MCP and A2A, layered onto workload identity you already run rather than a new identity system.

Why it matters

Pause it, roll it back, and prove what happened.

An agent that can act on your systems needs the same accountability you'd demand of a person with those permissions: an owner, approved limits, a record of what it did, and a way to stop and undo it. AskLedger makes that record signed and independently verifiable, so the answer to "what did the agent do, and who authorized it?" survives an audit.

AskLedger is at design-partner stage. The agent accountability workspace described here is on our roadmap; we're building it now with a small set of design partners.