Salmon
Request Demo

Execution Verification Infrastructure

Execution History for Autonomous Multi-Actor Software Systems

Salmon establishes cryptographically verifiable execution history and state lineage across humans, agents, and automation.

Explore the Protocol See How It Works

Execution flows forward.
Salmon goes upstream to the origin.

Salmon establishes verifiable execution history at the durable state boundary.

AI agents take actions.
Their execution leaves a trail.
Salmon makes it verifiable.

AI agents execute & delegate authority

Software Has Acquired Operational Agency

Autonomous software actors increasingly execute tools, modify files, invoke APIs, change infrastructure, and produce consequential software state.

  1. Human
  2. Agent
  3. Tool / API
  4. Runtime
  5. State change

The big shift

Software Has Become Multi-Actor

Humans are no longer the only consequential software actors. Agents, tools, runtimes, and autonomous control systems increasingly execute changes at machine speed.

When autonomous software changes production state, reconstructing what happened after the fact is often too late and incomplete.

The unique value of execution history

Resulting State ≠ Execution History

Identity
Tells us who is the actor.
Identity / Authority
Tells us what was permitted.
Observability
Tells us what was observed.
Control
Determines whether execution should proceed.
Execution Verification (Salmon)
Tells us what actually executed, and can that history be independently verified?

Different properties. Different questions. Execution Verification completes the picture.

The missing property

The Resulting State Does Not Preserve the Execution History that Produced It

Traditional observability provides useful signals, but not a verifiable record of how a state was produced.

Salmon establishes the record

A verifiable execution record with state lineage

Where it sits in the stack

Execution Verification Is a Distinct Infrastructure Layer

Salmon sits between consequential execution and the systems that consume execution evidence.

Learn about the architecture

  • Agents
  • Tools / APIs
  • Containers
  • Runtimes
  • Cloud / OS

Consequential execution

Execution Verification Infrastructure Capture · Execution Record · State Lineage · Verification

Verifiable execution evidence

  • AI / AgentsEvaluation · Supervision
  • SecurityInvestigation · Response
  • EngineeringDebugging · Reconstruction
  • Audit / GovernanceAccountability · Compliance

An example execution record

Execution History Is Established during Execution

Each execution event is cryptographically recorded, linked to state transitions and can be verified.

See how it works

Representative execution event

{
    "actor": "agent-a1a",
    "action": "write",
    "artifact": "src/auth/session.ts",
    "state_before": "91d7a4e2...",
    "state_after": "a4e8c1f9...",
    "predecessor": "04f7a2e1...",
    "timestamp": "2026-09-13T14:32:08Z",
    "signature": "ed25519:3a...",
    "status": "verified"
}
91d7a4e2 a4e8c1f9 S0 S1 EXEC
04F7A2 Verified

Explicit boundaries

What Wasn't Captured Remains Explicit

If execution continuity has not been established, Salmon does not visually or structurally manufacture it.

Machine-Speed Systems Require
Machine-Verifiable Execution History

Salmon. Execution Verification Infrastructure.

Explore the ProtocolSee How It Works