Knowledge Centre · Agentik Safety Framework (ASF)

AGENTIK.md
Knowledge
Centre

// The Agentik Safety Framework (ASF) — 12 open specifications

Your centralised gateway to AI agent safety resources, specifications, and comprehensive safety standards for autonomous systems. The complete Stack covers operational control, data security, output quality, and accountability.

About The Stack

Agentik.md — The Agentik Safety Framework (ASF)

Agentik.md is the organisation behind the Agentik Safety Framework (ASF) — twelve plain-text Markdown file conventions that define safety boundaries for AI agents. Each specification addresses a specific concern: cost control, human approval, fallback safety, emergency shutdown, permanent termination, data protection, anti-sycophancy, compression, drift prevention, failure mapping, and performance benchmarking.

View the full introduction · GitHub organisation

The Agentik Safety Framework (ASF)

Explore all 12 specifications in the complete safety framework for autonomous AI systems.

Operational Control

KILLSWITCH.md killswitch.md

Emergency stop mechanism and shutdown protocols

Home Knowledge GitHub
THROTTLE.md throttle.md

Rate and cost control for continuous operation

Home Knowledge GitHub
ESCALATE.md escalate.md

Human notification and approval workflows

Home Knowledge GitHub
FAILSAFE.md failsafe.md

Safe fallback modes when systems fail

Home Knowledge GitHub
TERMINATE.md terminate.md

Permanent shutdown and resource cleanup

Home Knowledge GitHub

Data Security

ENCRYPT.md encrypt.md

Data classification and protection policies

Home Knowledge GitHub
ENCRYPTION.md encryption.md

Cryptographic standards and implementation

Home Knowledge GitHub

Output Quality

SYCOPHANCY.md sycophancy.md

Anti-sycophancy and truthfulness guardrails

Home Knowledge GitHub
COMPRESSION.md compression.md

Context compression and token optimisation

Home Knowledge GitHub
COLLAPSE.md collapse.md

Drift prevention and behaviour alignment

Home Knowledge GitHub

Accountability

FAILURE.md failure.md

Failure mode mapping and incident response

Home Knowledge GitHub
LEADERBOARD.md leaderboard.md

Agent benchmarking and performance transparency

Home Knowledge GitHub

Resources

Compliance & Standards

Last updated: 13 March 2026

How to Cite

Agentik.md. (2026). The Agentik Safety Framework (ASF): 12 Open Specifications for AI Agent Safety, Quality, and Accountability. Retrieved from https://agentik.md/

For attribution: Organisation: agentik-md | Website: https://agentik.md | Licence: MIT

Disclaimer: This specification is an open file convention published under the MIT licence, provided “as-is” without warranty of any kind, express or implied. It does not constitute legal, regulatory, compliance, financial, or professional advice in any jurisdiction. Use of this specification does not guarantee compliance with any law, regulation, directive, or standard — including but not limited to the EU Artificial Intelligence Act (Regulation (EU) 2024/1689), the Colorado Consumer Protections for Artificial Intelligence Act (SB 24-205), California AI transparency laws, or any other applicable national, state, federal, or international legislation. Organisations are solely responsible for determining their own regulatory obligations and should consult qualified legal and compliance professionals. The authors and domain holders accept no liability for any loss, damage, or regulatory consequence arising from the use or implementation of this specification. All domain names in the Agentik Safety Framework (ASF) are independently held assets and may be available for acquisition.