CAIN-42 MCPGateAGENTIC AI 75 · 2026 → CAIN-42
Seventy-five 2026 agentic developments, governed
A CAIN-42 page for every major agentic-AI development of 2026: the authority it needs, the evidence that proves it stayed inside that authority, and the containment when it does not. Labels distinguish live controls from library code, simulations and designs.
Last reviewed 2026-11-11
75 developments, 7 moats, one control plane. The 2026 agentic-AI list describes capability. This library describes the authority, evidence and containment each capability needs before it should touch a consequential system. Every entry is labelled: 30 live, 24 library, 3 simulated, 18 proposed. Read the label before the claim.
How to read a page
Each page answers the same three questions: what authority does this capability need, what evidence proves the action stayed inside it, and what contains the action when the answer is unknown, wrong or hostile. The mapping is to CAIN-42's seven canonical moats. It is not a claim that the named vendor uses CAIN-42.
See all 75 in one table · Buy them as 75 products and services · machine-readable registry · sitemap · how the decision path works
Seven commercial bundles
The 75 pages package into seven buyer problems. The bundle is the sales unit; the page is the proof of specificity.
Code & Proof (12)
Prove the fix, not just the diff.
Platform engineering, AI coding-tool vendors, security engineering
$1,500–$6,000 per audit; $2,500–$4,000/month retainer
Streams 1, 2, 5 and 9 of the operator playbook
Memory & Context (8)
Memory that cannot launder authority.
Agent-platform teams, enterprise AI governance, regulated data owners
$2,500 per audit; metered from $0.0002 per inspection
Streams 2, 4 and 6 of the operator playbook
Identity & Authority (13)
Every agent, every delegation, every boundary.
CISO, IAM, platform security, agent-tool vendors
$2,500–$7,500 per review; integration from $10,000
Streams 2, 4 and 8 of the operator playbook
Commerce & Payments (13)
Delegated spending with a ceiling and a receipt.
Fintech, commerce platforms, marketplaces, treasury teams
$2,500–$7,500 per review; integration from $10,000
Streams 2, 4 and 9 of the operator playbook
Physical & Lab (9)
A validity check between the plan and the machine.
Lab automation, robotics, industrial operations, construction technology
$5,000–$15,000 per engagement
Streams 2 and 4 of the operator playbook; advisory, not a certified safety case
Swarm & Multi-Agent (5)
Delegation you can follow across every hop.
Agent orchestrator vendors, research labs, enterprise platform teams
$3,500–$7,500 per audit; $2,500–$4,000/month retainer
Streams 2, 3 and 5 of the operator playbook
Runtime Containment & Evidence (15)
Contain first, then prove what happened.
SOC, incident response, compliance, enterprise IT
$2,500–$5,000 per review; $2,000–$10,000 per evidence pack
Streams 2, 3 and 4 of the operator playbook
Code & Proof (12)
Prove the fix, not just the diff. Buyer: Platform engineering, AI coding-tool vendors, security engineering. Entry offer: Fabrication audit or long-run engineering attestation.
| # | 2026 development | CAIN-42 moat | Evidence | Offer |
|---|---|---|---|---|
| 1 | Full-day autonomous software engineering Software engineering | 4 · Execution Provenance & Attestation | LIVE | Long-Run Engineering Attestation |
| 4 | Machine-checkable security proofs Formal verification | 4 · Execution Provenance & Attestation | LIVE | Proof-Carrying Change Review |
| 9 | End-to-end autonomous R&D loops Research engineering | 7 · Autonomous Trust Control Loop | LIBRARY | Autonomous R&D Change Review |
| 10 | Failure-driven agent improvement Evaluation | 6 · Adversarial Trust Engine | LIVE | Failure-Driven Improvement Audit |
| 11 | Self-supervised theorem discovery Formal mathematics | 4 · Execution Provenance & Attestation | LIBRARY | Formal Artifact Review |
| 23 | Autonomous vulnerability investigation and repair Security engineering | 6 · Adversarial Trust Engine | LIVE | Vulnerability-to-Patch Proof |
| 24 | Coding agents inside native development tools Developer tooling | 4 · Execution Provenance & Attestation | LIVE | Coding-Agent Action Gate |
| 38 | Multi-agent desktop supervision Developer tooling | 4 · Execution Provenance & Attestation | LIVE | Parallel Agent Run Review |
| 41 | Multistep creative agents Creative production | 4 · Execution Provenance & Attestation | LIBRARY | Creative Workflow Provenance |
| 54 | Deployment-driven agent benchmarks Evaluation | 6 · Adversarial Trust Engine | LIVE | Deployment Trace Evaluation |
| 58 | Repeated-run comparisons of agent frameworks Evaluation | 6 · Adversarial Trust Engine | LIBRARY | Framework Evaluation Review |
| 62 | Local AI coding on personal computers Developer tooling | 6 · Adversarial Trust Engine | LIBRARY | Local Agent Containment Review |
Memory & Context (8)
Memory that cannot launder authority. Buyer: Agent-platform teams, enterprise AI governance, regulated data owners. Entry offer: Memory authority audit; hosted memory-sanitisation pilot.
| # | 2026 development | CAIN-42 moat | Evidence | Offer |
|---|---|---|---|---|
| 14 | Persistent, correction-aware agent memory Memory | 2 · Security-Context Continuity | LIVE | Memory Authority Audit |
| 15 | Retained reasoning and context compaction Context management | 4 · Execution Provenance & Attestation | LIBRARY | Context Integrity Review |
| 16 | Offline memory consolidation (dreaming) Memory | 2 · Security-Context Continuity | LIBRARY | Memory Consolidation Review |
| 29 | Personal context connected to mobile actions Mobile agents | 2 · Security-Context Continuity | LIBRARY | Personal Context Review |
| 31 | Shared organisational context graphs Knowledge graphs | 3 · Trust Graph | LIBRARY | Context Graph Review |
| 55 | Multi-session memory evaluation Memory evaluation | 2 · Security-Context Continuity | LIBRARY | Memory Evaluation Review |
| 65 | Work–personal data separation for mobile agents Mobile security | 2 · Security-Context Continuity | PROPOSED | Mobile Data Boundary Review |
| 72 | Reusable user-defined workflow skills Agent skills | 2 · Security-Context Continuity | LIVE | Saved Workflow Review |
Identity & Authority (13)
Every agent, every delegation, every boundary. Buyer: CISO, IAM, platform security, agent-tool vendors. Entry offer: Agent authority review; MCP/A2A boundary review.
| # | 2026 development | CAIN-42 moat | Evidence | Offer |
|---|---|---|---|---|
| 17 | Agent skills as a portable ecosystem standard Agent interoperability | 2 · Security-Context Continuity | LIVE | Portable Skill Review |
| 21 | Task-scoped, short-lived agent credentials Identity | 2 · Security-Context Continuity | LIVE | Agent Credential Review |
| 30 | Agent communication across business systems Enterprise integration | 2 · Security-Context Continuity | LIVE | Business-System Connector Review |
| 33 | Pathology agents as interoperable MCP tools Clinical tools | 2 · Security-Context Continuity | LIVE | Clinical Tool Boundary Review |
| 40 | Native creative-project editing through MCP Creative tools | 2 · Security-Context Continuity | LIVE | Creative Tool Boundary Review |
| 46 | Business ontologies for regulated-industry agents Enterprise knowledge | 3 · Trust Graph | LIBRARY | Ontology & Policy Review |
| 50 | Banking detection of agent activity Financial security | 1 · Trust State | LIBRARY | Agent Interaction Risk Review |
| 51 | Know Your Agent identity intelligence Identity | 1 · Trust State | LIBRARY | Agent Identity Review |
| 60 | Composable web-research toolchains Research tooling | 2 · Security-Context Continuity | LIVE | Research Toolchain Review |
| 61 | CRM-specialised reasoning models Domain models | 1 · Trust State | PROPOSED | Domain Model Authority Review |
| 63 | Agent permission review inside collaboration tools Workplace governance | 2 · Security-Context Continuity | LIVE | Chat Approval Review |
| 68 | Bring-your-own-agent advertising Advertising | 2 · Security-Context Continuity | PROPOSED | Third-Party Agent Review |
| 69 | Agentic advertising interoperability standards Standards | 2 · Security-Context Continuity | PROPOSED | Interop Authority Review |
Commerce & Payments (13)
Delegated spending with a ceiling and a receipt. Buyer: Fintech, commerce platforms, marketplaces, treasury teams. Entry offer: Payment-mandate review; browser-agent checkout review.
| # | 2026 development | CAIN-42 moat | Evidence | Offer |
|---|---|---|---|---|
| 22 | Bounded agent identities and wallets Payments | 1 · Trust State | LIVE | Agent Wallet Mandate Review |
| 28 | Cross-application mobile task execution Mobile agents | 5 · Predictive Trust & Blast Radius | LIVE | Mobile Agent Action Review |
| 47 | Universal commerce protocol Commerce | 2 · Security-Context Continuity | PROPOSED | Commerce Protocol Authority Review |
| 48 | Browser payments without exposing card details Payments | 2 · Security-Context Continuity | LIVE | Agent Checkout Mandate Review |
| 49 | Structured website actions through WebMCP Web agents | 2 · Security-Context Continuity | LIVE | Web Tool Definition Review |
| 52 | Controlled agent-to-agent commerce experiments Agent commerce | 2 · Security-Context Continuity | LIBRARY | Agent-to-Agent Deal Review |
| 53 | Consumer agents connected to real services Consumer agents | 2 · Security-Context Continuity | PROPOSED | Consumer Agent Authority Review |
| 57 | Shopping benchmarks that reward refusing bad purchases Evaluation | 1 · Trust State | LIVE | Commerce Refusal Review |
| 59 | More efficient browser representations and action loops Browser agents | 5 · Predictive Trust & Blast Radius | LIVE | Browser Agent Control Review |
| 66 | Mobile browser agents for real-world errands Browser agents | 5 · Predictive Trust & Blast Radius | LIVE | Browser Errand Review |
| 67 | Agent-to-agent advertising execution Advertising | 2 · Security-Context Continuity | PROPOSED | Campaign Authority Review |
| 70 | Retailer-controlled agents in search Commerce | 1 · Trust State | PROPOSED | Brand Agent Review |
| 74 | Broader international mobile-agent deployment Mobile agents | 2 · Security-Context Continuity | PROPOSED | Mobile Expansion Review |
Physical & Lab (9)
A validity check between the plan and the machine. Buyer: Lab automation, robotics, industrial operations, construction technology. Entry offer: Physical-action assurance review (advisory, simulation-first).
| # | 2026 development | CAIN-42 moat | Evidence | Offer |
|---|---|---|---|---|
| 2 | Agent-operated molecular-biology laboratories Scientific labs | 5 · Predictive Trust & Blast Radius | SIMULATED | Physical Autonomy Assurance Review |
| 3 | Closed-loop experimental discovery Scientific discovery | 7 · Autonomous Trust Control Loop | LIBRARY | Discovery Loop Evidence Review |
| 7 | Digital twins for safe physical agent actions Physical safety | 5 · Predictive Trust & Blast Radius | SIMULATED | Physical Action Twin Review |
| 8 | Robot programming through agent teams Robotics | 4 · Execution Provenance & Attestation | SIMULATED | Robotics Release Gate Review |
| 12 | Experimentally validated biological binder discovery Biotechnology | 4 · Execution Provenance & Attestation | PROPOSED | Research Evidence Review |
| 32 | Agent-assisted search through biological archives Scientific data | 4 · Execution Provenance & Attestation | PROPOSED | Research Retrieval Provenance |
| 43 | Agentic construction scheduling Construction | 5 · Predictive Trust & Blast Radius | PROPOSED | Physical Schedule Assurance Review |
| 44 | Evidence-based construction change orders Construction | 4 · Execution Provenance & Attestation | PROPOSED | Contract Evidence Review |
| 73 | Industrial expertise captured as agents Industrial knowledge | 3 · Trust Graph | PROPOSED | Industrial Knowledge Review |
Swarm & Multi-Agent (5)
Delegation you can follow across every hop. Buyer: Agent orchestrator vendors, research labs, enterprise platform teams. Entry offer: Swarm delegation audit; collusion and drift review.
| # | 2026 development | CAIN-42 moat | Evidence | Offer |
|---|---|---|---|---|
| 5 | Open-world multi-agent mathematical discovery Mathematical research | 3 · Trust Graph | LIBRARY | Research Provenance Review |
| 6 | Single-prompt coordination of specialised agent teams Agent orchestration | 2 · Security-Context Continuity | LIVE | Swarm Delegation Audit |
| 13 | Multi-model workflow orchestration Agent orchestration | 1 · Trust State | LIBRARY | Multi-Model Orchestration Review |
| 25 | Reported orchestration of 10,000 concurrent agents Scale | 3 · Trust Graph | PROPOSED | Fleet Governance Design Review |
| 39 | Faster models for collaborative subagents Model infrastructure | 5 · Predictive Trust & Blast Radius | LIBRARY | Fast-Path Authority Review |
Runtime Containment & Evidence (15)
Contain first, then prove what happened. Buyer: SOC, incident response, compliance, enterprise IT. Entry offer: Runtime containment review; signed evidence pack.
| # | 2026 development | CAIN-42 moat | Evidence | Offer |
|---|---|---|---|---|
| 18 | Full-stack agent safety infrastructure Platform security | 7 · Autonomous Trust Control Loop | LIVE | Enterprise Agent Governance Review |
| 19 | Runtime agent quarantine Runtime containment | 7 · Autonomous Trust Control Loop | LIVE | Runtime Containment Audit |
| 20 | Operating-system-enforced agent containment Endpoint security | 6 · Adversarial Trust Engine | LIBRARY | OS Containment Review |
| 26 | Finished knowledge-work deliverables Knowledge work | 4 · Execution Provenance & Attestation | LIVE | Signed Deliverable Pack |
| 27 | Always-on workplace teammates Workplace agents | 7 · Autonomous Trust Control Loop | LIBRARY | Agent Lease Review |
| 34 | Clinical-trial oversight agents Clinical operations | 4 · Execution Provenance & Attestation | PROPOSED | Regulated Workflow Evidence Review |
| 35 | Voice agents that reason while listening Voice interfaces | 5 · Predictive Trust & Blast Radius | PROPOSED | Real-Time Agent Gate Review |
| 36 | Evidence-grounded threat investigation Security operations | 4 · Execution Provenance & Attestation | LIBRARY | Threat Investigation Evidence Review |
| 37 | Enterprise agent execution with governance Enterprise platform | 7 · Autonomous Trust Control Loop | LIVE | Enterprise Execution Governance |
| 42 | Continuous creative-production agents Creative operations | 7 · Autonomous Trust Control Loop | PROPOSED | Background Agent Lease Review |
| 45 | IT investigation with approval-gated remediation IT operations | 5 · Predictive Trust & Blast Radius | LIVE | Approval-Gated Remediation Review |
| 56 | Visually complex knowledge-work evaluation Evaluation | 4 · Execution Provenance & Attestation | LIBRARY | Benchmark Evidence Review |
| 64 | Enterprise agent inventory and cost accountability Agent operations | 1 · Trust State | LIBRARY | Agent Inventory Review |
| 71 | Proactive personal assistants Personal agents | 7 · Autonomous Trust Control Loop | LIBRARY | Proactive Agent Lease Review |
| 75 | Observational evidence of larger task delegation Adoption research | 4 · Execution Provenance & Attestation | LIVE | Delegation Measurement Review |
What this page is not. CAIN-42 is not affiliated with, endorsed by, integrated into, or shipped with the organisation named in the 2026 item. The item is public context; the page describes how CAIN-42 would govern that class of capability. No page here claims a customer, a partnership, a certification, a patent, or a production integration. Prices are proposals. Built by a three-person team.