Is Agenomic's replay deterministic?

No. Replay is statistical: it compares output distributions and behavior-contract outcomes across re-runs rather than asserting bit-identical results.

Does Agenomic help with EU AI Act compliance?

Yes. Each release can produce an EU AI Act Articles 9–15 mapping plus signed evidence you can attach to technical documentation and record-keeping. Agenomic provides evidence; it is not a legal certification.

Do I need an account or the cloud to use Agenomic?

No. The CLI and SDKs run fully offline with no account. Cloud is optional — for sharing signed bundles, reviews, and replay.

What does an agent genome capture?

Prompts, models and configuration, tools, permissions, policies, dependencies, memory contracts, execution traces, behavioral fingerprints, and compliance evidence.

How is this different from observability or AgentOps?

Observability tells you what your agent did; Agenomic tells you what your agent is — a signed, versioned, replayable record reviewers can diff before promotion.

Are Agenomic's audit artifacts tamper-evident?

Yes. Genomes are signed and produce tamper-evident attestations (ATEP) suitable for compliance review and post-incident analysis.

Compliance

Build an audit trail for AI agents (EU AI Act Article 12)

Article 12 of the EU AI Act requires automatic record-keeping over a system’s lifetime. Here is how Agenomic’s signed traces, attestations, and replay give AI agents a tamper-evident audit trail.

Lire en français Search index Edit on GitHub

What Article 12 requires

Article 12 asks high-risk AI systems to automatically record events (“logs”) over their lifetime, so the system functioning is traceable and post-incident review is possible. For an AI agent, the events that matter are the ones that explain why it did what it did.

Automatic recording of events throughout the system lifecycle.
Traceability appropriate to the system’s intended purpose.
Records that support post-market monitoring and incident review.

What to record for an agent

An agent’s behavior is more than its final answer. Agenomic captures the execution trace — the sequence of model calls, tool calls, inputs, and outputs — alongside the genome that defines the agent, so a record stays meaningful months later.

Inputs and outputs for each run.
Model calls (provider, model, configuration) and tool calls.
The genome version in force: prompts, policies, permissions, and dependencies.
Behavioral fingerprints used to compare runs.

How Agenomic makes the trail tamper-evident

Capturing logs is not enough if they can be edited after the fact. Agenomic signs genomes and emits tamper-evident attestations (ATEP), so an archived record can be verified rather than trusted. Traces export to JSONL for your own storage and pipelines.

Signed bundles and attestations make after-the-fact edits detectable.
Traces export to traces.jsonl for offline storage.
Agenomic Cloud (optional) can collect signed bundles into evidence packs with approvals.

Replay as living evidence

A static log shows what happened once. Replay re-runs an agent and compares output distributions and behavior-contract outcomes against a baseline, so your audit trail also shows whether behavior held over time. Replay is statistical — it does not assert bit-for-bit identical results.

Get started

Follow the quickstart to capture your first trace, then archive the signed bundle from each release. Agenomic provides this evidence; it does not issue legal certification, and Article 12 sign-off remains your responsibility.