# AGENTIC 75 products: evidence, 2026-11-12

**75 productised offers. 49 engine-backed and verified today. Every status published.**

| Status | Count | Meaning |
|---|---:|---|
| DELIVERABLE | 49 | An engine in the CAIN-42 repo produced a bounded artifact from a sample and its manifest verified. |
| PILOT | 16 | Library code or a live mechanism exists; a bounded pilot needs engineering time. |
| ADVISORY | 0 | Simulation or advisory only. No production engine. |
| DESIGN | 10 | Design-partner engagement. No control delivered. |

Public surfaces:

- Product catalog: https://cainstudio.online/agentic-products
- Governance map: https://cainstudio.online/agentic-75
- Product catalog JSON: https://cainstudio.online/agentic-products.json
- Developer home: https://mcpgate.online/agentic-75
- Hugging Face Space: https://huggingface.co/spaces/vicentenoriega/CAIN-42

## What was run

`python3 -m cain_agentic_products.cli selftest` ran every DELIVERABLE SKU and re-verified each
manifest. Result: **49/49 verified**. The full sample deliveries included here are:

- `CAIN-A75-001` Long-Run Engineering Attestation (`cain_a75_engines/longrun/engine.py`)
- `CAIN-A75-002` Physical Autonomy Assurance Review (`cain_a75_engines/labpreflight/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-003` Discovery Loop Evidence Review (`cain_a75_engines/discoveryloop/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-004` Proof-Carrying Change Review (`cain_a75_engines/proofreview/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-005` Research Provenance Review (`cain_a75_engines/researchgraph/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-007` Physical Action Twin Review (`cain_a75_engines/physicaltwin/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-008` Robotics Release Gate Review (`cain_a75_engines/roboticsgate/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-009` Autonomous R&D Change Review (`cain_a75_engines/rndchange/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-010` Failure-Driven Improvement Audit (`cain_a75_engines/failurefix/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-011` Formal Artifact Review (`cain_a75_engines/formalartifact/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-012` Research Evidence Review (`cain_a75_engines/binderevidence/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-013` Multi-Model Orchestration Review (`cain_a75_engines/modelorchestration/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-014` Memory Authority Audit (`cain/memory.py`)
- `CAIN-A75-015` Context Integrity Review (`cain_a75_engines/contextcompaction/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-016` Memory Consolidation Review (`cain_a75_engines/memoryconsolidation/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-018` Enterprise Agent Governance Review (`cain_a75_engines/enterprisegov/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-019` Runtime Containment Audit (`cain_a75_engines/runtimecontainment/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-020` OS Containment Review (`cain_a75_engines/oscontainment/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-024` Coding-Agent Action Gate (`cain_a75_engines/codinggate/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-025` Fleet Governance Design Review (`cain_a75_engines/fleetdesign/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-026` Signed Deliverable Pack (`cain/worm_export.py`)
- `CAIN-A75-027` Agent Lease Review (`cain_a75_engines/agentlease/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-028` Mobile Agent Action Review (`cain_a75_engines/mobilegate/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-029` Personal Context Review (`cain_a75_engines/personalcontext/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-031` Context Graph Review (`cain_a75_engines/contextgraph/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-032` Research Retrieval Provenance (`cain_a75_engines/retrievalprov/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-034` Regulated Workflow Evidence Review (`cain_a75_engines/regulatedevidence/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-035` Real-Time Agent Gate Review (`cain_a75_engines/realtimegate/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-036` Threat Investigation Evidence Review (`cain_a75_engines/threatevidence/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-037` Enterprise Execution Governance (`cain_a75_engines/executiongov/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-038` Parallel Agent Run Review (`cain_a75_engines/parallelruns/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-039` Fast-Path Authority Review (`cain_a75_engines/fastpath/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-041` Creative Workflow Provenance (`cain_a75_engines/creativeprov/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-042` Background Agent Lease Review (`cain_a75_engines/backgroundlease/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-043` Physical Schedule Assurance Review (`cain_a75_engines/constructionschedule/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-044` Contract Evidence Review (`cain_a75_engines/contractevidence/engine.py`) · signed run + standalone `verify_offline.py` in `signed-sample/`
- `CAIN-A75-057` Commerce Refusal Review (`cain/commerce/mandate_guard.py`)

## Files

| File | What it is |
|---|---|
| `CATALOG.json` | all 75 SKUs, statuses, prices, engines, development links |
| `SELFTEST.json` | the 49 engine runs, each manifest hash, and verification result |
| `INVARIANTS.json` | 10 checks over the catalog and the deliveries |
| `deliveries/` | 37 complete sample deliveries (`findings.json`, `report.md`, `manifest.json`; signed runs in `signed-sample/`) |
| `verify_agentic75.py` | re-check everything above; imports no CAIN code, stdlib only |
| `LIMITATIONS.md` | what this bundle does not show |
| `MANIFEST.json` | SHA-256 of every file here |

## Check it yourself

```bash
python3 verify_agentic75.py
python3 -c "import hashlib,json;m=json.load(open('MANIFEST.json'));print(all(hashlib.sha256(open(f,'rb').read()).hexdigest()==v['sha256'] for f,v in m['files'].items()))"
```

## What this does not show

This is **not** evidence that all 75 offers are finished engines. It is evidence that 75 offers exist,
49 run real engines, and every status and limit is published. See `LIMITATIONS.md`.