Skip to content

Aevum

The Python library that makes your AI agent's memory accountable.

Aevum is a Python library that gives AI agents a signed audit trail, consent-checked data access, and verifiable decision records — three problems that tend to surface together in production. The quickstart gets you to working code in ten minutes.

Developer

New here? Start with AEVUM_DEV=1 — Quickstart →

Compliance

Evaluating for EU AI Act Article 12? — Article 12 guide →

Security

Reviewing the threat model? — THREAT_MODEL.md →

Where to start

  • Understand it first

    How Aevum works: the governed membrane, the sigchain, the five barriers, and the consent model — in one page.

    Architecture

  • Build with it

    First governed session in under 10 minutes. Works on Linux, macOS, and Windows.

    Quickstart

  • Evaluate it

    Threat model, security architecture, and common security questions for engineers evaluating production deployment.

    Security

Install

pip install aevum-core

Apache-2.0. No telemetry. Runs fully offline.

Self-hosted — your data never leaves your infrastructure. No vendor API. No licensing server. No SaaS dependency.

Who It Is For

Aevum is useful when your AI agent needs to answer any of these questions:

  • "What did the agent know when it made this decision?" — replay()
  • "Did the agent have permission to access this data?" — consent ledger
  • "Has this audit trail been tampered with?" — verify_sigchain()
  • "Was a human in the loop for this action?" — review()

Good fit: Regulated industries (healthcare, finance, legal), agentic workflows that touch PII or sensitive data, compliance-driven environments (EU AI Act, HIPAA, SOC 2), and teams that need to demonstrate audit accountability to customers or regulators.

Not a fit yet: If you need a streaming data platform, an AI orchestration framework, a database, or an agent execution runtime — Aevum is not those things. See What Aevum Is Not for the normative list.

See It in Action

Aevum governs its own development using its own governed functions. The self-governance pipeline (aevum-labs/aevum-maintainer) demonstrates what an Aevum-governed workflow looks like end-to-end.

Self-governance pipeline status:

Phase Deliverable Status
1 Scaffold + OIDC ingest + Cedar policies ✅ Complete
2 MCP research interface (6 read-only tools) ✅ Complete
3 Structured consent gate (HITL approval + dwell time) ✅ Complete
4 Replay endpoint + Rekor anchor + break-glass ✅ Complete
5 Demo page ✅ Complete — demo.aevum.build

Read how it works: How Aevum Governs Itself