# Methodology & Bibliography — Agentic Glossary

> How the Agentic Glossary is built, what counts as a canonical citation, and how we keep the entries fresh. Maintained by AgentsBooks.

**Data last refreshed:** 2026-05-07. **Next scheduled refresh:** Q3 2026 (or sooner on regulator / vendor trigger).

## Sourcing rule

Every glossary entry has at least one direct quote from a canonical primary source — meaning:

- The vendor's own published documentation, blog, or research paper (Anthropic, Google, OpenAI, Microsoft, Meta).
- The standards body's own specification (Linux Foundation A2A / MCP, NIST, ISO).
- The regulator's own publication (European Commission, NIST, FATF, FCA).
- The peer-reviewed paper of record (arXiv-published papers cited >500 times, NeurIPS / ICLR proceedings).
- An analyst publication where the analyst is the primary source for the framing (Gartner Hype Cycle terminology, McKinsey adoption stats).

Wikipedia, secondary blogs, and content farms are *never* the primary citation.

## Freshness flags

- **Foundational** — citations to original peer-reviewed papers (e.g., 2020 RAG, 2022 Constitutional AI). Considered evergreen.
- **In force** — current binding regulator text or in-effect specifications. Includes the in-force date.
- **Emerging 2026** — terms that entered mainstream discourse in 2026 (ADLC, AX, context graph, agent management platform). Quarterly review.
- **Contested** — entries with named, meaningful disagreement in the field. We carry both positions.

Anything older than six months that does not carry one of the four flags is considered stale and gets retired or refreshed.

## Refresh cadence

| Trigger | Action |
|---|---|
| Quarterly | Full audit: every cited URL pinged, every primary source re-read for material changes |
| Vendor canonical-source publication | Targeted refresh of affected entries within 7 days |
| Regulator publication or in-force date change | Same-week update |
| New peer-reviewed paper that supersedes a cited claim | Same-week update |
| URL 404 or vendor pivot | Immediate fix |

## Bibliography (v0 — 2026-05-07)

1. Anthropic, *Building Effective Agents* — https://www.anthropic.com/research/building-effective-agents — accessed 2026-05-07
2. Anthropic, *Measuring AI agent autonomy in practice* — https://www.anthropic.com/research/measuring-agent-autonomy — accessed 2026-05-07
3. Bai et al. (Anthropic), *Constitutional AI: Harmlessness from AI Feedback*, arXiv:2212.08073, 2022 — https://arxiv.org/abs/2212.08073 — accessed 2026-05-07 [Foundational]
4. Lewis et al., *Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks*, arXiv:2005.11401, 2020 — https://arxiv.org/abs/2005.11401 — accessed 2026-05-07 [Foundational]
5. Anthropic, *Introducing the Model Context Protocol* — https://www.anthropic.com/news/model-context-protocol — accessed 2026-05-07
6. Model Context Protocol, *Specification (2025-11-25)* — https://modelcontextprotocol.io/specification/2025-11-25 — accessed 2026-05-07
7. Google Developers, *Announcing the Agent2Agent Protocol (A2A)* — https://developers.googleblog.com/en/a2a-a-new-era-of-agent-interoperability/ — accessed 2026-05-07
8. A2A Protocol, *Specification* — https://a2a-protocol.org/latest/specification/ — accessed 2026-05-07 [In force]
9. NIST, *Artificial Intelligence Risk Management Framework (AI 100-1)* — https://nvlpubs.nist.gov/nistpubs/ai/nist.ai.100-1.pdf — accessed 2026-05-07 [In force]
10. NIST, *AI Risk Management Framework hub* (AI Agent Standards Initiative announced February 2026 via CAISI) — https://www.nist.gov/itl/ai-risk-management-framework — accessed 2026-05-07
11. European Commission, *EU AI Act Implementation Timeline* — https://ai-act-service-desk.ec.europa.eu/en/ai-act/timeline/timeline-implementation-eu-ai-act — accessed 2026-05-07 [In force 2 August 2026, subject to Digital Omnibus deferral proposal]
12. Gartner, *Hype Cycle for Agentic AI 2026* — https://www.gartner.com/en/articles/hype-cycle-for-agentic-ai — accessed 2026-05-07
13. McKinsey QuantumBlack, *The state of AI in 2025: Agents, innovation, and transformation*, November 2025 — https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai — accessed 2026-05-07
14. Marcus, G., *Six (or seven) predictions for AI 2026 from a Generative AI realist*, Substack — https://garymarcus.substack.com/p/six-or-seven-predictions-for-ai-2026 — accessed 2026-05-07 [Contested]
15. AgentsBooks, *Anatomy of a Firm* — https://agentsbooks.com/anatomy — accessed 2026-05-07

## What counts as a "term"?

The bar for inclusion in v0:

1. It is searched for, currently, by builders / operators / buyers of AI agents (validated via search-intent signals).
2. At least one canonical primary source defines it directly.
3. It connects to one of the eight AgentsBooks primitives or to a current pillar topic (P1–P8).

v0 ships with 31 terms across six categories. v1 will expand toward 100+ as new vocabulary stabilises.

## How to suggest a term or correction

Email hello@agentsbooks.com with the term, what you'd like it defined as, and the canonical source you want quoted. We aim to triage within five business days.

---

*Updated 2026-05-07.*