Skip to content
ai-contributors logomark ai-contributor-spec / audit
Specification › Levels overview

Five levels of AI autonomy.

The ladder

Section titled “The ladder”

The level names, descriptions, workflow summaries, and rule counts below come from AI-CONTRIBUTOR-RULE-CATALOG.json. Each level adds gates on top of the previous one.

L0Baseline Hygiene
AI is not part of the contribution workflow yet. People may still use AI for personal help, but no AI tool reads the repository as workspace context, creates commits, or opens pull requests.0% autonomy

The repository satisfies the baseline requirements in §1, §2, and §5: no secrets in version control, placeholder-only env examples when contributor-supplied or runtime environment variables exist, documented credential handling, a clean setup path, committed lockfile, pinned runtime, pinned package manager, and automated formatting. AI is not yet part of the contribution workflow. A repository at Level 0 publishes `conformance_level: 0`.

8 catalog rows start here
AI Contributor L0 Baseline Hygiene
L1Hardened
AI tools may read repository context, explain code, suggest commands, or help with review, but do not produce shippable changes25% autonomy

Level 0, plus the remaining unconditional `MUST` rows in Pillars 1–4 and the oversight MUSTs in §23, §24, and §25. A Level 1 repository is strong enough for AI tooling to read repository content. AI-specific MUSTs in Pillars 5–6 are not yet evaluated.

60 catalog rows start here
AI Contributor L1 Hardened
L2AI Assisted
AI may produce code, docs, tests, configuration, or migrations that a human actively accepts. Extra risks such as tool use, external GitHub content, fetched URLs, retained context, provider routing, and dependency suggestions must be controlled.50% autonomy

Level 1, plus AI instructions, AI data classification, AI-surface redaction, AI-generated-content licensing, and authorship tracking. Any triggered §18–§22 or §26 rows must also be satisfied. AI may assist humans, but every change passes through human acceptance.

37 catalog rows start here
AI Contributor L2 AI Assisted
L3AI Authored
AI may complete delegated implementation tasks, open pull requests, or make material repository changes for human review75% autonomy

Level 2, plus the rows triggered by AI-authored work, shared skills, MCP servers, delegated agents, or other AI tooling that materially authors code. `AIC-prompt-audit-trail` always applies because Level 3 means AI materially authors code that ships. Humans still review before merge.

14 catalog rows start here
AI Contributor L3 AI Authored
L4AI Autonomous
AI may merge, release, deploy, schedule recurring changes, approve workflows, rotate settings, or otherwise act without per-change human approval100% autonomy

Level 3, plus any autonomous-runner rows and every applicable `SHOULD`. Each `SHOULD` row must be `Fulfilled` or `Not relevant` with a documented reason. `⚠️ Warning` is not passing for Level 4. This level is for repositories where AI ships changes without human review for each change.

81 catalog rows start here
AI Contributor L4 AI Autonomous

Display your level

Section titled “Display your level”

Once the audit shows every row at your target level passing, advertise the level via a badge in your README. The labels below are derived from the catalog levels. Use the badge for the highest level your current audit actually reaches. The pinned form is recommended because it ties the claim to the catalog version you audited against.

<!-- README.md -->
[![AI Contributor: L2 AI Assisted](https://img.shields.io/badge/AI%20Contributor-L2%20AI%20Assisted-green)](https://ai-contributors.dev/ai-contributor-spec/docs/conformance-levels/)

<!-- Pinned to v0.1.2 -->
[![AI Contributor: L2 AI Assisted · spec v0.1.2](https://img.shields.io/badge/AI%20Contributor-L2%20AI%20Assisted-green?rev=v0.1.2)](https://github.com/ai-contributors/ai-contributor-spec/releases/tag/v0.1.2)

How a level is claimed

Section titled “How a level is claimed”

A level is claimed by running the audit and showing every row at your target level passing. The audit records the spec source, audited commit, level claimed, and evidence hash so the claim can be reproduced.

Claims are unilateral: there is no central registry. The verification path is to reproduce the audit against the repository at a specific commit and catalog version.

Where to next

Section titled “Where to next”