BotConduct Standard (BCS)

The behavioral governance layer for bots and AI agents.

Identity tells you WHO a bot is. Conduct tells you HOW it behaves.

BCS measures the HOW.

What is BCS?

BCS is a proprietary behavioral scoring engine that measures bot and AI agent conduct under evolving conditions. The engine evaluates trajectory — how conduct changes as signals, state, and adversarial inputs change — not single-snapshot compliance against a static checklist. It produces a verdict-only score (0–100) and rating that platforms can use to govern automated traffic with granularity.

How it works

BCS observes bot behavior passively across its network of integrated sites. Each bot is scored on a proprietary rubric covering identity, compliance, resource usage, and operational transparency. The methodology, weights, and thresholds are trade secrets — only the verdict is exposed.

Output:

Score: 0–100 (continuous)

Rating: `exemplary` | `acceptable` | `non_compliant` | `hostile`

Action recommendation: `allow` | `throttle` | `challenge` | `block`

For Platforms

BCS is designed as infrastructure-level middleware. Platforms integrate BCS at the edge or origin to govern bot traffic across all their hosted apps — without requiring action from individual site owners.

[Integration details →](/for-sites.html)

For Bot Operators

Operators can certify their bots at three levels (all free):

**Level 1:** Self-declared identity (30 seconds)

**Level 2:** Email-verified + active behavioral test (90-day cert)

**Level 3:** Ed25519 cryptographic identity ("Fast Pass") — portable trust across all BCS-integrated sites

[Level 3 spec →](/spec/level-3) | [Get certified →](/get-certified.html)

Registry

BCS maintains a public registry of 172+ observed bot operators with aggregate scores.

GET https://botconduct.org/api/lookup/{BotName}

[View registry →](/api/registry)

FAQ

[Frequently asked questions →](/faq.html)

botconduct.org — hello@botconduct.org