
reference
com.aigovernanceengineer/reference
Read-only AI governance reference: obligations, crosswalks, controls, glossary, patterns, chapters.
Documentation
AI Governance Engineering: The Thesis & Body of Knowledge
AI governance engineering is the application of engineering practice (systems thinking, product thinking and code) to the governance of AI systems.
This repository is the founding text of that discipline: a short Thesis and a growing body of knowledge. It is written for practitioners who build governance for AI systems (policy-as- code, agent registries, eval gates, runtime guardrails, machine-readable evidence) and who measure their work by realised risk reduction and audit-ready evidence, not by framework coverage.
It is version 0.5.0: 24 chapters, 13 of them new (11 to 23, from AI defined for governance through the programme, the lifecycle and the law to governing agents), 33 patterns with their own pages, an open-data API, a browser toolkit, and open control profiles and research notes (draft), still deliberately incomplete and open to contributions. The Body of Knowledge, website, datasets and project materials are authored by Jorge García Aibar. The Thesis is the sole exception: it is co-authored by Jorge García Aibar and Aurélie Pols and is open for signatures.
Home: https://aigovernanceengineer.com (the Thesis at /thesis, Body of Knowledge at /bok).
Source repository: github.com/losanchos5/aige.
Start here
- THESIS.md: the founding statement (definition, the five problems with legacy AI governance, eight values, six principles, what AI governance engineers build).
- bok/01-definition.md: what the discipline is, and the eight neighbours it is not.
- OUTLINE.md: the full table of contents with a brief for each chapter.
Structure
aige/
├── THESIS.md The founding statement (one page)
├── STYLEGUIDE.md Voice, formatting, templates, citation format, terminology
├── OUTLINE.md Full table of contents with per-chapter briefs
├── README.md This file
├── bok/ The Body of Knowledge
│ ├── 00-preface.md Provenance, audience, how to cite, versioning
│ ├── 01-definition.md Definition, disambiguation, object of governance, three questions
│ ├── 02-why-now.md The five problems with the evidence; market, regulatory and agent signals
│ ├── 03-values-principles.md The eight values and six principles, expanded
│ ├── 04-the-stack.md The five-layer reference architecture
│ ├── 05-patterns.md Pattern catalogue (CSIRO template): one summary per pattern
│ ├── patterns/ One file per pattern (bok/patterns/<slug>.md, /patterns/<slug>)
│ ├── 06-the-role.md The AI governance engineer, by workflow, skills and market
│ ├── 07-maturity-model.md Five levels from Documented to Continuous
│ ├── 08-regulatory-map.md Obligation → artefact → layer reverse index
│ ├── 09-glossary.md Canonical definitions, alphabetical
│ ├── 10-reading-list.md Annotated bibliography
│ ├── 11-…23-*.md AI defined, the programme, the lifecycle, the law, governing agents
│ ├── CONTRIBUTORS.md Authorship, contributors, SIGNATORIES
│ └── CHANGELOG.md Version history
├── research/ Technical notes (research/<slug>.md, served at /research/<slug>)
├── sources/
│ └── SOURCES.md Consolidated, verified source table (one section per chapter)
└── build/
└── build_pdf.py Assembles the Markdown into dist/site-preview.html (PDF export optional)
Chapters 00–23 are drafted; contributions that extend the catalogue, mappings and arguments are welcome and credited.
The five-layer stack
The artefacts an AI governance engineer ships map onto five layers, always named in this order:
- Govern-as-Code: policy-as-code, framework crosswalks, gates in CI/CD.
- Inventory & Transparency: agent registry, AIBOM, model and data cards, FRIA/DPIA.
- Evals & Red Teaming as Evidence: adversarial and capability evals wired into the pipeline.
- Runtime Controls & Observability: guardrails, agent identity, tracing, kill switches.
- Assurance & Continuous Compliance: machine-readable evidence (OSCAL), continuous assurance, serious-incident reporting.
Open datasets
The site publishes two reusable datasets, generated from the same source as its reference pages:
| Dataset | JSON | CSV | Versioned source |
|---|---|---|---|
| Framework crosswalk | JSON | CSV | crosswalk.ts |
| Obligation → artefact → layer mappings | JSON | CSV | frameworks.ts |
These are curated reference mappings. They are illustrative, not a claim of conformity or a substitute for reading the underlying instruments. The live exports can change as the site evolves; record the retrieval date and source commit when using them in research.
How to cite the datasets
Cite the parent work, name the dataset you used, and include its URL and retrieval date:
García Aibar, J. AI Governance Engineering: The Thesis & Body of Knowledge (v0.5.0). https://doi.org/10.5281/zenodo.22956197. Dataset: Framework crosswalk / Obligation mappings. [Dataset URL; retrieval date; source commit].
The DOI identifies the archived parent release, not a separately deposited dataset. The DOI above is the v0.5.0 version DOI; the concept DOI resolves to the latest archived release. See CITATION.cff for citation metadata. Reuse is covered by this repository’s licence.
How to contribute
See CONTRIBUTING.md for the full contribution guide. In short:
- Read STYLEGUIDE.md and follow the chapter or pattern template exactly.
- Source every factual claim with a
[n]citation, tag itprimary/secondary/reported, and add the row tosources/SOURCES.mdunder your chapter's section. - Open a pull request. Keep the voice practitioner-direct; name categories, not vendors; never invent a statistic.
To review a control profile or a research note, propose a failure mode, share an implementation example or map a control to a clause, use the GitHub issue forms listed at https://aigovernanceengineer.com/contribute.
How to sign
Sign the Thesis by opening a pull request that adds your name (and optionally a LinkedIn or GitHub link) to the SIGNATORIES section of bok/CONTRIBUTORS.md.
Licence
Except where otherwise noted, this work is licensed under CC BY 4.0. You may share and adapt it provided you give appropriate credit, link to the licence and indicate changes. Attribute the Body of Knowledge, website, datasets and project materials to Jorge García Aibar. Attribute the Thesis to Jorge García Aibar and Aurélie Pols.
No installation packages available.