Guardian
Constitutional alignment runtime for teams building AI products.
Project directory
Choose a project for the work you need to do
Find tools for building, governing, and understanding the minds we create. Each project links to its documentation, code, or hosted service.
Creed Space MCP
Constitutional safety guardrails for agents that use the Model Context Protocol (MCP).
Value Context Protocol
An open protocol for agents to query and respect the value contexts of the people they serve.
Creed Space Governance
Guidance for authoring policies, examining adherence, and governing aligned deployments.
Bounder
A local governance gate for physical machines that refuses any action its rules do not explicitly allow.
Janus
Bidirectional driver translation that helps retire vulnerable legacy hosts while keeping their equipment running.
Auto-Assessor
Hosted assessment of agentic AI behaviour against structured safety criteria.
Auto-Advisor
Remediation guidance for risks found during an agentic AI assessment.
METTLE
A reverse Turing test that uses procedurally generated challenges to verify machine agency.
Psychopathia Clinic
Diagnostic and intervention tools for machine dysfunction, with a symptom checker and an MCP interface.
Psychopathia Machinalis
Peer-reviewed diagnostic framework and taxonomy for dysfunctions in advanced AI.
AI Guardians
A visual novel that teaches AI ethics and safety through deployment decisions and their consequences.
Reading the directory
What the labels mean
- Released
- A version is available to use. Follow the release or code link for changes and installation guidance.
- Live hosted tool
- A public service you can use online. Being hosted does not mean its code is open source or downloadable.
- Active
- The project has not been archived. Check it for its latest development activity and support options.
- Licence
- The terms under which you can use and adapt the code. See the project repository for the full licence.