acf-mcp 1.0.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/LICENSE +21 -0
- package/README.md +87 -0
- package/content/fiches/.gitkeep +0 -0
- package/content/fiches/ACF-00.en.md +59 -0
- package/content/fiches/ACF-00.fr.md +59 -0
- package/content/fiches/ACF-01.en.md +32 -0
- package/content/fiches/ACF-01.fr.md +32 -0
- package/content/fiches/ACF-02.en.md +32 -0
- package/content/fiches/ACF-02.fr.md +32 -0
- package/content/fiches/ACF-03.en.md +78 -0
- package/content/fiches/ACF-03.fr.md +78 -0
- package/content/fiches/ACF-04.en.md +32 -0
- package/content/fiches/ACF-04.fr.md +32 -0
- package/content/fiches/ACF-05.en.md +32 -0
- package/content/fiches/ACF-05.fr.md +32 -0
- package/content/fiches/ACF-06.en.md +32 -0
- package/content/fiches/ACF-06.fr.md +32 -0
- package/content/fiches/ACF-07.en.md +32 -0
- package/content/fiches/ACF-07.fr.md +32 -0
- package/content/fiches/ACF-08.en.md +32 -0
- package/content/fiches/ACF-08.fr.md +32 -0
- package/content/fiches/ACF-09.en.md +32 -0
- package/content/fiches/ACF-09.fr.md +32 -0
- package/content/fiches/ACF-10.en.md +32 -0
- package/content/fiches/ACF-10.fr.md +32 -0
- package/content/fiches/ACF-11.en.md +32 -0
- package/content/fiches/ACF-11.fr.md +32 -0
- package/content/fiches/ACF-12.en.md +32 -0
- package/content/fiches/ACF-12.fr.md +32 -0
- package/content/fiches/ACF-13.en.md +32 -0
- package/content/fiches/ACF-13.fr.md +32 -0
- package/content/fiches/ACF-14.en.md +32 -0
- package/content/fiches/ACF-14.fr.md +32 -0
- package/content/fiches/ACF-15.en.md +32 -0
- package/content/fiches/ACF-15.fr.md +32 -0
- package/content/fiches/ACF-16.en.md +32 -0
- package/content/fiches/ACF-16.fr.md +32 -0
- package/content/framework/.gitkeep +0 -0
- package/content/framework/autonomy-levels.json +73 -0
- package/content/framework/ddao.json +16 -0
- package/content/framework/dimensions.json +83 -0
- package/content/framework/principles.json +77 -0
- package/content/glossary/.gitkeep +0 -0
- package/content/glossary/en.json +96 -0
- package/content/glossary/fr.json +96 -0
- package/content/guides/.gitkeep +0 -0
- package/content/manual/.gitkeep +0 -0
- package/content/meta.json +23 -0
- package/content/rules/.gitkeep +0 -0
- package/content/rules/ai-act-annex-i.json +54 -0
- package/content/rules/ai-act-annex-iii.json +165 -0
- package/content/rules/ai-act-roles.json +52 -0
- package/content/rules/autonomy-inference.json +38 -0
- package/content/rules/criticality-matrix.json +83 -0
- package/content/rules/ddao-controls-mapping.json +148 -0
- package/content/rules/gdpr-qualification.json +58 -0
- package/content/rules/gpai-triggers.json +41 -0
- package/content/rules/rules-meta.json +5 -0
- package/content/rules/sign-off-matrix.json +84 -0
- package/content/whitepaper/.gitkeep +0 -0
- package/dist/index.d.ts +30 -0
- package/dist/index.js +2898 -0
- package/dist/index.js.map +1 -0
- package/dist/lib/rate-limit.d.ts +24 -0
- package/dist/lib/rate-limit.js +89 -0
- package/dist/lib/rate-limit.js.map +1 -0
- package/dist/search-index.json +1 -0
- package/dist/server-DxztThU5.d.ts +931 -0
- package/dist/transport/http.d.ts +27 -0
- package/dist/transport/http.js +2908 -0
- package/dist/transport/http.js.map +1 -0
- package/dist/transport/stdio.d.ts +1 -0
- package/dist/transport/stdio.js +2907 -0
- package/dist/transport/stdio.js.map +1 -0
- package/package.json +81 -0
package/LICENSE
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
1
|
+
MIT License
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2026 Vincent Dorange / ACF Standard
|
|
4
|
+
|
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
|
7
|
+
in the Software without restriction, including without limitation the rights
|
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
|
10
|
+
furnished to do so, subject to the following conditions:
|
|
11
|
+
|
|
12
|
+
The above copyright notice and this permission notice shall be included in all
|
|
13
|
+
copies or substantial portions of the Software.
|
|
14
|
+
|
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
21
|
+
SOFTWARE.
|
package/README.md
ADDED
|
@@ -0,0 +1,87 @@
|
|
|
1
|
+
# acf-mcp
|
|
2
|
+
|
|
3
|
+
**Official ACF® Model Context Protocol server.**
|
|
4
|
+
Agentic Commerce Framework® — first open governance standard for autonomous AI agents.
|
|
5
|
+
|
|
6
|
+
[ACF Standard](https://acfstandard.com) · [Compliance](https://acfstandard.com/compliance) · [Teaching toolkit](https://acfstandard.com/enseignants)
|
|
7
|
+
|
|
8
|
+
---
|
|
9
|
+
|
|
10
|
+
## What it does
|
|
11
|
+
|
|
12
|
+
`acf-mcp` exposes the ACF® doctrine — 4 principles, 4 autonomy levels, 6 maturity dimensions, the DDAO role, 17 methodological cards, 5 expert regulatory guides (AI Act, GDPR, DORA, NIS2, ISO 42001), the white paper and the glossary — as native MCP **Resources**, **Tools** and **Prompts** so any MCP client (Claude Desktop, Cursor, Windsurf, Continue…) can reason on top of it.
|
|
13
|
+
|
|
14
|
+
It also ships **7 deterministic REASON tools** (built on a versioned knowledge base — no internal LLM call) that take a user case and return a structured ACF® governance assessment:
|
|
15
|
+
|
|
16
|
+
- `acf.advisor` — generic case → structured advice (autonomy + risk + principles + fiches + first actions)
|
|
17
|
+
- `acf.classify-agent` — **killer tool**: 10-field qualified-enum input → preliminary qualification (ACF level + criticality + AI Act role + GDPR status + obligations by lifecycle + DDAO controls + sign-off requirements)
|
|
18
|
+
- `acf.assess-autonomy` — N0-N3 recommendation with go/no-go + gating + kill switch design
|
|
19
|
+
- `acf.identify-governance-gaps` — 6-dimension maturity score + prioritised remediation
|
|
20
|
+
- `acf.map-ai-act-obligations` — full obligation set structured by lifecycle phase (pre-go-live / continuous / on-incident) with Digital Omnibus deferrals
|
|
21
|
+
- `acf.assign-ddao-controls` — level × risk → recommended + ACF-canonical controls
|
|
22
|
+
- `acf.evaluate-agent-mandate` — 8-check audit of an existing mandate
|
|
23
|
+
|
|
24
|
+
Every REASON output is **signed**: `doctrine_version`, `doctrine_hash`, `doctrine_archive_url`, `regulatory_snapshot`, `generated_at`. Every REASON output is positioned as **preliminary qualification, not legal advice** — `requires_human_review: true` is constant in V1.0.
|
|
25
|
+
|
|
26
|
+
## Install (stdio)
|
|
27
|
+
|
|
28
|
+
```bash
|
|
29
|
+
# Claude Desktop, Cursor, Windsurf:
|
|
30
|
+
npx -y acf-mcp
|
|
31
|
+
```
|
|
32
|
+
|
|
33
|
+
Add to Claude Desktop `claude_desktop_config.json`:
|
|
34
|
+
|
|
35
|
+
```json
|
|
36
|
+
{
|
|
37
|
+
"mcpServers": {
|
|
38
|
+
"acf": {
|
|
39
|
+
"type": "stdio",
|
|
40
|
+
"command": "npx",
|
|
41
|
+
"args": ["-y", "acf-mcp"]
|
|
42
|
+
}
|
|
43
|
+
}
|
|
44
|
+
}
|
|
45
|
+
```
|
|
46
|
+
|
|
47
|
+
## Use over HTTP (remote)
|
|
48
|
+
|
|
49
|
+
```json
|
|
50
|
+
{
|
|
51
|
+
"mcpServers": {
|
|
52
|
+
"acf": { "type": "http", "url": "https://mcp.acfstandard.com" }
|
|
53
|
+
}
|
|
54
|
+
}
|
|
55
|
+
```
|
|
56
|
+
|
|
57
|
+
No API key — V1.0 is open access (cf. `Position stratégique` in the spec).
|
|
58
|
+
|
|
59
|
+
## Locales
|
|
60
|
+
|
|
61
|
+
V1.0 ships FR + EN translated. The 11 other locales (`es, de, pt, it, nl, ru, ar, tr, ja, zh, ko`) are infrastructure-supported but fallback to EN with `_meta.is_fallback: true` signalled. Full translation comes in V1.3 with native juristic review per jurisdiction.
|
|
62
|
+
|
|
63
|
+
## Doctrine traceability
|
|
64
|
+
|
|
65
|
+
Every REASON output embeds:
|
|
66
|
+
- `doctrine_version` — e.g. `"ACF framework v1.0 / rules 2026-06"`
|
|
67
|
+
- `doctrine_hash` — sha256 of the active content + rules at call time
|
|
68
|
+
- `doctrine_archive_url` — permanent archive of the doctrine version used
|
|
69
|
+
- `regulatory_snapshot` — corpus that was reasoned against
|
|
70
|
+
- `generated_at` — ISO 8601 timestamp
|
|
71
|
+
|
|
72
|
+
An audit can reconstruct the exact analysis produced at instant T by loading the archive URL.
|
|
73
|
+
|
|
74
|
+
## License
|
|
75
|
+
|
|
76
|
+
MIT. See `LICENSE`.
|
|
77
|
+
|
|
78
|
+
## Citation
|
|
79
|
+
|
|
80
|
+
```
|
|
81
|
+
Dorange, V. (2026). Agentic Commerce Framework® (ACF®). ACF Standard.
|
|
82
|
+
https://acfstandard.com
|
|
83
|
+
```
|
|
84
|
+
|
|
85
|
+
## Contributing
|
|
86
|
+
|
|
87
|
+
Issues + PRs welcome on https://github.com/aiconsulting06000-tech/acfstandard. Doctrine modifications go through the maintainer review process.
|
|
File without changes
|
|
@@ -0,0 +1,59 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-00"
|
|
3
|
+
slug: "introduction-acf"
|
|
4
|
+
title: "Introduction to ACF®"
|
|
5
|
+
title_en: "Introduction to ACF®"
|
|
6
|
+
order: 0
|
|
7
|
+
maturity_dimension: "D2"
|
|
8
|
+
related_principles: ["P1", "P2", "P3", "P4"]
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: ["ACF-01", "ACF-03"]
|
|
11
|
+
keywords: ["framework", "introduction", "governance", "agentic", "doctrine"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-00_introduction.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Introduction to ACF®
|
|
17
|
+
|
|
18
|
+
## Purpose of this card
|
|
19
|
+
|
|
20
|
+
This card sets the overall context of the Agentic Commerce Framework® (ACF®). It is the doctrinal entry point for anyone discovering the methodology: DDAO, DPO, CISO, lawyer, compliance officer, business sponsor, teacher, student.
|
|
21
|
+
|
|
22
|
+
## Why ACF® exists
|
|
23
|
+
|
|
24
|
+
Agentic AI — agents that take and execute decisions without a human in the loop at every step — raises a question that traditional governance frameworks did not address: who answers when the agent acts?
|
|
25
|
+
|
|
26
|
+
ACF® formalises the answer: the organisation. Always. And it provides a methodological apparatus to make that accountability operable.
|
|
27
|
+
|
|
28
|
+
## The 4 founding principles
|
|
29
|
+
|
|
30
|
+
1. **Decision Sovereignty** (P1) — accountability is never delegated.
|
|
31
|
+
2. **Doctrinal Traceability** (P2) — every decision is reconstructible.
|
|
32
|
+
3. **Ultimate Human Control** (P3) — the kill switch stays active.
|
|
33
|
+
4. **Proportional Governance** (P4) — control intensity follows criticality.
|
|
34
|
+
|
|
35
|
+
## The 4 autonomy levels
|
|
36
|
+
|
|
37
|
+
- **N0** Assistance — the agent suggests, the human decides.
|
|
38
|
+
- **N1** Supervised recommendation — the human validates.
|
|
39
|
+
- **N2** Conditional execution — the agent decides within a bounded perimeter.
|
|
40
|
+
- **N3** Autonomous execution — the agent decides and executes without prior validation.
|
|
41
|
+
|
|
42
|
+
## The 6 maturity dimensions
|
|
43
|
+
|
|
44
|
+
D1 Strategy & governance · D2 Doctrine & framework · D3 Design & technical control · D4 Accountability & roles · D5 Regulatory compliance · D6 Audit & continuous improvement.
|
|
45
|
+
|
|
46
|
+
## The DDAO role
|
|
47
|
+
|
|
48
|
+
Delegated Decision Agent Officer. A guardian role defined by ACF® to carry operational accountability for one or more agents (cf. DDAO card in `framework/ddao`).
|
|
49
|
+
|
|
50
|
+
## Recommended reading
|
|
51
|
+
|
|
52
|
+
- ACF-01 (Autonomy mapping)
|
|
53
|
+
- ACF-03 (Agentic Constitution)
|
|
54
|
+
- ACF-12 (Agent Mandate)
|
|
55
|
+
- The ACF® white paper for the extended doctrine.
|
|
56
|
+
|
|
57
|
+
## Teaching use
|
|
58
|
+
|
|
59
|
+
ACF® is freely usable in classes, seminars and continuing education. The 17 cards are calibrated teaching materials. See `/enseignants` on acfstandard.com for the full toolkit.
|
|
@@ -0,0 +1,59 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-00"
|
|
3
|
+
slug: "introduction-acf"
|
|
4
|
+
title: "Introduction à l'ACF®"
|
|
5
|
+
title_en: "Introduction to ACF®"
|
|
6
|
+
order: 0
|
|
7
|
+
maturity_dimension: "D2"
|
|
8
|
+
related_principles: ["P1", "P2", "P3", "P4"]
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: ["ACF-01", "ACF-03"]
|
|
11
|
+
keywords: ["framework", "introduction", "gouvernance", "agentique", "doctrine"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-00_introduction.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Introduction à l'ACF®
|
|
17
|
+
|
|
18
|
+
## Objet de la fiche
|
|
19
|
+
|
|
20
|
+
Cette fiche pose le cadre général de l'Agentic Commerce Framework® (ACF®). Elle est le point d'entrée doctrinal pour toute personne qui découvre la méthodologie : DDAO, DPO, RSSI, juriste, compliance officer, business sponsor, enseignant, étudiant.
|
|
21
|
+
|
|
22
|
+
## Pourquoi ACF® existe
|
|
23
|
+
|
|
24
|
+
L'IA agentique — des agents qui prennent et exécutent des décisions sans humain dans la boucle à chaque étape — pose une question que les frameworks de gouvernance traditionnels ne traitaient pas : qui répond quand l'agent agit ?
|
|
25
|
+
|
|
26
|
+
ACF® formalise la réponse : l'organisation. Toujours. Et il fournit un appareil méthodologique pour rendre cette responsabilité opérable.
|
|
27
|
+
|
|
28
|
+
## Les 4 principes fondateurs
|
|
29
|
+
|
|
30
|
+
1. **Souveraineté décisionnelle** (P1) — la responsabilité ne se délègue jamais.
|
|
31
|
+
2. **Traçabilité doctrinale** (P2) — toute décision est reconstituable.
|
|
32
|
+
3. **Contrôle humain ultime** (P3) — le kill switch reste actif.
|
|
33
|
+
4. **Proportionnalité de la gouvernance** (P4) — l'intensité du contrôle suit la criticité.
|
|
34
|
+
|
|
35
|
+
## Les 4 niveaux d'autonomie
|
|
36
|
+
|
|
37
|
+
- **N0** Assistance — l'agent suggère, l'humain décide.
|
|
38
|
+
- **N1** Recommandation supervisée — l'humain valide.
|
|
39
|
+
- **N2** Exécution conditionnelle — l'agent décide dans un périmètre borné.
|
|
40
|
+
- **N3** Exécution autonome — l'agent décide et exécute sans validation préalable.
|
|
41
|
+
|
|
42
|
+
## Les 6 dimensions de maturité
|
|
43
|
+
|
|
44
|
+
D1 Stratégie & gouvernance · D2 Doctrine & cadre · D3 Conception & contrôle technique · D4 Responsabilités & rôles · D5 Conformité réglementaire · D6 Audit & amélioration continue.
|
|
45
|
+
|
|
46
|
+
## Le rôle DDAO
|
|
47
|
+
|
|
48
|
+
Delegated Decision Agent Officer. Rôle gardien défini par ACF® pour porter la responsabilité opérationnelle d'un ou plusieurs agents (cf. fiche DDAO dans `framework/ddao`).
|
|
49
|
+
|
|
50
|
+
## Lecture conseillée
|
|
51
|
+
|
|
52
|
+
- ACF-01 (Cartographie de l'autonomie)
|
|
53
|
+
- ACF-03 (Constitution Agentique)
|
|
54
|
+
- ACF-12 (Mandat d'Agent)
|
|
55
|
+
- Le livre blanc ACF® pour la doctrine étendue.
|
|
56
|
+
|
|
57
|
+
## Cas d'usage pédagogique
|
|
58
|
+
|
|
59
|
+
ACF® est librement utilisable en cours, séminaire et formation continue. Les 17 fiches sont des supports d'enseignement calibrés. Voir `/enseignants` sur acfstandard.com pour le toolkit complet.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-01"
|
|
3
|
+
slug: "cartographie-autonomie"
|
|
4
|
+
title: "Autonomy Mapping"
|
|
5
|
+
title_en: "Autonomy Mapping"
|
|
6
|
+
order: 1
|
|
7
|
+
maturity_dimension: "D2"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["mapping", "autonomy", "levels"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-01_cartographie-autonomie.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Autonomy Mapping
|
|
17
|
+
|
|
18
|
+
## Purpose of this card
|
|
19
|
+
|
|
20
|
+
To complete — extracted from the ACF® toolkit PDF `/toolkit/ACF-01_cartographie-autonomie.pdf`. This card is part of the canonical set of 17 ACF® methodological cards.
|
|
21
|
+
|
|
22
|
+
## Methodology
|
|
23
|
+
|
|
24
|
+
To complete.
|
|
25
|
+
|
|
26
|
+
## Example
|
|
27
|
+
|
|
28
|
+
To complete.
|
|
29
|
+
|
|
30
|
+
## Links to other cards
|
|
31
|
+
|
|
32
|
+
To complete.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-01"
|
|
3
|
+
slug: "cartographie-autonomie"
|
|
4
|
+
title: "Cartographie de l'autonomie"
|
|
5
|
+
title_en: "Autonomy Mapping"
|
|
6
|
+
order: 1
|
|
7
|
+
maturity_dimension: "D2"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["cartographie", "autonomie", "niveaux"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-01_cartographie-autonomie.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Cartographie de l'autonomie
|
|
17
|
+
|
|
18
|
+
## Objet de la fiche
|
|
19
|
+
|
|
20
|
+
À compléter — extraction depuis le PDF du toolkit ACF® `/toolkit/ACF-01_cartographie-autonomie.pdf`. Cette fiche fait partie du jeu canonique des 17 cartes méthodologiques ACF®.
|
|
21
|
+
|
|
22
|
+
## Méthodologie
|
|
23
|
+
|
|
24
|
+
À compléter.
|
|
25
|
+
|
|
26
|
+
## Exemple
|
|
27
|
+
|
|
28
|
+
À compléter.
|
|
29
|
+
|
|
30
|
+
## Liens vers d'autres fiches
|
|
31
|
+
|
|
32
|
+
À compléter.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-02"
|
|
3
|
+
slug: "matrice-criticite"
|
|
4
|
+
title: "Criticality Matrix"
|
|
5
|
+
title_en: "Criticality Matrix"
|
|
6
|
+
order: 2
|
|
7
|
+
maturity_dimension: "D5"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["criticality", "matrix", "calibration"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-02_matrice-criticite.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Criticality Matrix
|
|
17
|
+
|
|
18
|
+
## Purpose of this card
|
|
19
|
+
|
|
20
|
+
To complete — extracted from the ACF® toolkit PDF `/toolkit/ACF-02_matrice-criticite.pdf`. This card is part of the canonical set of 17 ACF® methodological cards.
|
|
21
|
+
|
|
22
|
+
## Methodology
|
|
23
|
+
|
|
24
|
+
To complete.
|
|
25
|
+
|
|
26
|
+
## Example
|
|
27
|
+
|
|
28
|
+
To complete.
|
|
29
|
+
|
|
30
|
+
## Links to other cards
|
|
31
|
+
|
|
32
|
+
To complete.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-02"
|
|
3
|
+
slug: "matrice-criticite"
|
|
4
|
+
title: "Matrice de Criticité"
|
|
5
|
+
title_en: "Criticality Matrix"
|
|
6
|
+
order: 2
|
|
7
|
+
maturity_dimension: "D5"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["criticité", "matrice", "calibration"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-02_matrice-criticite.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Matrice de Criticité
|
|
17
|
+
|
|
18
|
+
## Objet de la fiche
|
|
19
|
+
|
|
20
|
+
À compléter — extraction depuis le PDF du toolkit ACF® `/toolkit/ACF-02_matrice-criticite.pdf`. Cette fiche fait partie du jeu canonique des 17 cartes méthodologiques ACF®.
|
|
21
|
+
|
|
22
|
+
## Méthodologie
|
|
23
|
+
|
|
24
|
+
À compléter.
|
|
25
|
+
|
|
26
|
+
## Exemple
|
|
27
|
+
|
|
28
|
+
À compléter.
|
|
29
|
+
|
|
30
|
+
## Liens vers d'autres fiches
|
|
31
|
+
|
|
32
|
+
À compléter.
|
|
@@ -0,0 +1,78 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-03"
|
|
3
|
+
slug: "agentic-constitution"
|
|
4
|
+
title: "Agentic Constitution"
|
|
5
|
+
title_en: "Agentic Constitution"
|
|
6
|
+
order: 3
|
|
7
|
+
maturity_dimension: "D2"
|
|
8
|
+
related_principles: ["P1", "P3"]
|
|
9
|
+
related_articles:
|
|
10
|
+
- { regulation: "ai-act", article: "9", paragraph: "1" }
|
|
11
|
+
- { regulation: "gdpr", article: "25" }
|
|
12
|
+
related_fiches: ["ACF-04", "ACF-12"]
|
|
13
|
+
keywords: ["mandate", "perimeter", "agent", "constitution", "doctrine"]
|
|
14
|
+
version: "1.0"
|
|
15
|
+
pdf_url: "/toolkit/ACF-03_constitution-agentique.pdf"
|
|
16
|
+
---
|
|
17
|
+
|
|
18
|
+
# Agentic Constitution
|
|
19
|
+
|
|
20
|
+
## Purpose of this card
|
|
21
|
+
|
|
22
|
+
Define the founding document of an autonomous AI agent — its "constitution" — that sets the perimeter, operating principles, limits and governance commitments before any deployment.
|
|
23
|
+
|
|
24
|
+
## Methodology
|
|
25
|
+
|
|
26
|
+
### 1. Identification
|
|
27
|
+
|
|
28
|
+
- Agent name.
|
|
29
|
+
- ACF® doctrine version applied.
|
|
30
|
+
- Target autonomy level (N0 → N3).
|
|
31
|
+
- Appointed DDAO.
|
|
32
|
+
|
|
33
|
+
### 2. Decision perimeter
|
|
34
|
+
|
|
35
|
+
- Allowed decisions (positive, limitative list).
|
|
36
|
+
- Forbidden decisions (explicit negative list).
|
|
37
|
+
- Invocation boundaries (who can activate, from which channels).
|
|
38
|
+
|
|
39
|
+
### 3. Governance commitments
|
|
40
|
+
|
|
41
|
+
- Adherence to operational mandate (cf. ACF-12).
|
|
42
|
+
- Adherence to escalation thresholds (cf. ACF-09).
|
|
43
|
+
- Adherence to the kill switch (cf. ACF-07).
|
|
44
|
+
- Maintenance of the decision register (cf. ACF-05).
|
|
45
|
+
|
|
46
|
+
### 4. Reference doctrine
|
|
47
|
+
|
|
48
|
+
- Citation of the ACF® version applied (`doctrine_version`, `doctrine_hash`, `doctrine_archive_url`).
|
|
49
|
+
- Activated principles named (P1-P4).
|
|
50
|
+
- Mobilised cards named.
|
|
51
|
+
|
|
52
|
+
### 5. Validation
|
|
53
|
+
|
|
54
|
+
- DDAO sign-off mandatory.
|
|
55
|
+
- Sign-off from concerned functions (DPO if personal data, CISO if attack surface, Legal if legal exposure).
|
|
56
|
+
- Scheduled review date.
|
|
57
|
+
|
|
58
|
+
## "Lendari" example
|
|
59
|
+
|
|
60
|
+
Lendari deploys a stock-replenishment agent across 7 stores. The agentic constitution specifies:
|
|
61
|
+
- N2, replenishment perimeter bounded by price and quantity windows.
|
|
62
|
+
- DDAO = Supply Chain Director.
|
|
63
|
+
- Allowed: place orders within window; alert on imminent stockout.
|
|
64
|
+
- Forbidden: modify supplier panel; negotiate terms; commit payments.
|
|
65
|
+
- Kill switch: immediate suspension by DDAO or by the executive committee.
|
|
66
|
+
- Mandate: quarterly refresh, annual review.
|
|
67
|
+
|
|
68
|
+
## Links to other cards
|
|
69
|
+
|
|
70
|
+
- ACF-04 (Agent Card)
|
|
71
|
+
- ACF-12 (Agent Mandate)
|
|
72
|
+
- ACF-07 (Kill switch)
|
|
73
|
+
|
|
74
|
+
## Frequent mistakes
|
|
75
|
+
|
|
76
|
+
- Vague constitution that does not enumerate forbidden decisions → the agent drifts as soon as a case is not explicitly covered.
|
|
77
|
+
- DDAO appointed in title, not in practice → no one carries the decision.
|
|
78
|
+
- No doctrine version cited → impossible to trace after the fact.
|
|
@@ -0,0 +1,78 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-03"
|
|
3
|
+
slug: "constitution-agentique"
|
|
4
|
+
title: "Constitution Agentique"
|
|
5
|
+
title_en: "Agentic Constitution"
|
|
6
|
+
order: 3
|
|
7
|
+
maturity_dimension: "D2"
|
|
8
|
+
related_principles: ["P1", "P3"]
|
|
9
|
+
related_articles:
|
|
10
|
+
- { regulation: "ai-act", article: "9", paragraph: "1" }
|
|
11
|
+
- { regulation: "gdpr", article: "25" }
|
|
12
|
+
related_fiches: ["ACF-04", "ACF-12"]
|
|
13
|
+
keywords: ["mandat", "périmètre", "agent", "constitution", "doctrine"]
|
|
14
|
+
version: "1.0"
|
|
15
|
+
pdf_url: "/toolkit/ACF-03_constitution-agentique.pdf"
|
|
16
|
+
---
|
|
17
|
+
|
|
18
|
+
# Constitution Agentique
|
|
19
|
+
|
|
20
|
+
## Objet de la fiche
|
|
21
|
+
|
|
22
|
+
Définir le document fondateur d'un agent IA autonome — sa "constitution" — qui pose le périmètre, les principes opératoires, les limites et les engagements de gouvernance avant tout déploiement.
|
|
23
|
+
|
|
24
|
+
## Méthodologie
|
|
25
|
+
|
|
26
|
+
### 1. Identification
|
|
27
|
+
|
|
28
|
+
- Nom de l'agent.
|
|
29
|
+
- Version de la doctrine ACF® appliquée.
|
|
30
|
+
- Niveau d'autonomie cible (N0 → N3).
|
|
31
|
+
- DDAO nommé.
|
|
32
|
+
|
|
33
|
+
### 2. Périmètre de décision
|
|
34
|
+
|
|
35
|
+
- Décisions autorisées (liste positive et limitative).
|
|
36
|
+
- Décisions interdites (liste négative explicite).
|
|
37
|
+
- Frontières d'invocation (qui peut activer l'agent, depuis quels canaux).
|
|
38
|
+
|
|
39
|
+
### 3. Engagements de gouvernance
|
|
40
|
+
|
|
41
|
+
- Respect du mandat opérationnel (cf. ACF-12).
|
|
42
|
+
- Respect des seuils d'escalade (cf. ACF-09).
|
|
43
|
+
- Respect du kill switch (cf. ACF-07).
|
|
44
|
+
- Tenue du registre des décisions (cf. ACF-05).
|
|
45
|
+
|
|
46
|
+
### 4. Doctrine de référence
|
|
47
|
+
|
|
48
|
+
- Citation de la version ACF® appliquée (`doctrine_version`, `doctrine_hash`, `doctrine_archive_url`).
|
|
49
|
+
- Mention des principes activés (P1-P4).
|
|
50
|
+
- Mention des fiches mobilisées.
|
|
51
|
+
|
|
52
|
+
### 5. Validation
|
|
53
|
+
|
|
54
|
+
- Sign-off DDAO obligatoire.
|
|
55
|
+
- Sign-off des fonctions concernées (DPO si données personnelles, RSSI si surface d'attaque, Legal si exposition juridique).
|
|
56
|
+
- Date de revue programmée.
|
|
57
|
+
|
|
58
|
+
## Exemple « Lendari »
|
|
59
|
+
|
|
60
|
+
Lendari déploie un agent de réordonnancement de stock dans 7 boutiques. La constitution agentique précise :
|
|
61
|
+
- N2, périmètre de réordonnancement borné par la fourchette de prix et de quantité.
|
|
62
|
+
- DDAO = Directrice supply chain.
|
|
63
|
+
- Décisions autorisées : passer commande dans la fourchette ; alerter sur rupture imminente.
|
|
64
|
+
- Décisions interdites : modifier le panel fournisseurs ; négocier des conditions ; engager des paiements.
|
|
65
|
+
- Kill switch : suspension immédiate par la DDAO ou par le COMEX.
|
|
66
|
+
- Mandat : refresh trimestriel, revue annuelle.
|
|
67
|
+
|
|
68
|
+
## Liens vers d'autres fiches
|
|
69
|
+
|
|
70
|
+
- ACF-04 (Fiche Agent)
|
|
71
|
+
- ACF-12 (Mandat d'Agent)
|
|
72
|
+
- ACF-07 (Kill switch)
|
|
73
|
+
|
|
74
|
+
## Erreurs fréquentes
|
|
75
|
+
|
|
76
|
+
- Constitution vague qui n'énumère pas les décisions interdites → l'agent dérive dès que le cas n'est pas explicitement couvert.
|
|
77
|
+
- DDAO nommé en titre, pas en pratique → personne ne porte la décision.
|
|
78
|
+
- Pas de version de doctrine citée → impossible de tracer ex post.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-04"
|
|
3
|
+
slug: "fiche-agent"
|
|
4
|
+
title: "Agent Card"
|
|
5
|
+
title_en: "Agent Card"
|
|
6
|
+
order: 4
|
|
7
|
+
maturity_dimension: "D4"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["agent", "card", "identification"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-04_fiche-agent.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Agent Card
|
|
17
|
+
|
|
18
|
+
## Purpose of this card
|
|
19
|
+
|
|
20
|
+
To complete — extracted from the ACF® toolkit PDF `/toolkit/ACF-04_fiche-agent.pdf`. This card is part of the canonical set of 17 ACF® methodological cards.
|
|
21
|
+
|
|
22
|
+
## Methodology
|
|
23
|
+
|
|
24
|
+
To complete.
|
|
25
|
+
|
|
26
|
+
## Example
|
|
27
|
+
|
|
28
|
+
To complete.
|
|
29
|
+
|
|
30
|
+
## Links to other cards
|
|
31
|
+
|
|
32
|
+
To complete.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-04"
|
|
3
|
+
slug: "fiche-agent"
|
|
4
|
+
title: "Fiche Agent"
|
|
5
|
+
title_en: "Agent Card"
|
|
6
|
+
order: 4
|
|
7
|
+
maturity_dimension: "D4"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["agent", "fiche", "identification"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-04_fiche-agent.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Fiche Agent
|
|
17
|
+
|
|
18
|
+
## Objet de la fiche
|
|
19
|
+
|
|
20
|
+
À compléter — extraction depuis le PDF du toolkit ACF® `/toolkit/ACF-04_fiche-agent.pdf`. Cette fiche fait partie du jeu canonique des 17 cartes méthodologiques ACF®.
|
|
21
|
+
|
|
22
|
+
## Méthodologie
|
|
23
|
+
|
|
24
|
+
À compléter.
|
|
25
|
+
|
|
26
|
+
## Exemple
|
|
27
|
+
|
|
28
|
+
À compléter.
|
|
29
|
+
|
|
30
|
+
## Liens vers d'autres fiches
|
|
31
|
+
|
|
32
|
+
À compléter.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-05"
|
|
3
|
+
slug: "registre-decisions"
|
|
4
|
+
title: "Decision Register"
|
|
5
|
+
title_en: "Decision Register"
|
|
6
|
+
order: 5
|
|
7
|
+
maturity_dimension: "D2"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["register", "decisions", "traceability"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-05_registre-decisions.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Decision Register
|
|
17
|
+
|
|
18
|
+
## Purpose of this card
|
|
19
|
+
|
|
20
|
+
To complete — extracted from the ACF® toolkit PDF `/toolkit/ACF-05_registre-decisions.pdf`. This card is part of the canonical set of 17 ACF® methodological cards.
|
|
21
|
+
|
|
22
|
+
## Methodology
|
|
23
|
+
|
|
24
|
+
To complete.
|
|
25
|
+
|
|
26
|
+
## Example
|
|
27
|
+
|
|
28
|
+
To complete.
|
|
29
|
+
|
|
30
|
+
## Links to other cards
|
|
31
|
+
|
|
32
|
+
To complete.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-05"
|
|
3
|
+
slug: "registre-decisions"
|
|
4
|
+
title: "Registre des décisions"
|
|
5
|
+
title_en: "Decision Register"
|
|
6
|
+
order: 5
|
|
7
|
+
maturity_dimension: "D2"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["registre", "décisions", "traçabilité"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-05_registre-decisions.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Registre des décisions
|
|
17
|
+
|
|
18
|
+
## Objet de la fiche
|
|
19
|
+
|
|
20
|
+
À compléter — extraction depuis le PDF du toolkit ACF® `/toolkit/ACF-05_registre-decisions.pdf`. Cette fiche fait partie du jeu canonique des 17 cartes méthodologiques ACF®.
|
|
21
|
+
|
|
22
|
+
## Méthodologie
|
|
23
|
+
|
|
24
|
+
À compléter.
|
|
25
|
+
|
|
26
|
+
## Exemple
|
|
27
|
+
|
|
28
|
+
À compléter.
|
|
29
|
+
|
|
30
|
+
## Liens vers d'autres fiches
|
|
31
|
+
|
|
32
|
+
À compléter.
|
|
@@ -0,0 +1,32 @@
|
|
|
1
|
+
---
|
|
2
|
+
code: "ACF-06"
|
|
3
|
+
slug: "constitution-ecosysteme"
|
|
4
|
+
title: "Ecosystem Constitution"
|
|
5
|
+
title_en: "Ecosystem Constitution"
|
|
6
|
+
order: 6
|
|
7
|
+
maturity_dimension: "D1"
|
|
8
|
+
related_principles: []
|
|
9
|
+
related_articles: []
|
|
10
|
+
related_fiches: []
|
|
11
|
+
keywords: ["ecosystem", "constitution", "governance"]
|
|
12
|
+
version: "1.0"
|
|
13
|
+
pdf_url: "/toolkit/ACF-06_constitution-ecosysteme.pdf"
|
|
14
|
+
---
|
|
15
|
+
|
|
16
|
+
# Ecosystem Constitution
|
|
17
|
+
|
|
18
|
+
## Purpose of this card
|
|
19
|
+
|
|
20
|
+
To complete — extracted from the ACF® toolkit PDF `/toolkit/ACF-06_constitution-ecosysteme.pdf`. This card is part of the canonical set of 17 ACF® methodological cards.
|
|
21
|
+
|
|
22
|
+
## Methodology
|
|
23
|
+
|
|
24
|
+
To complete.
|
|
25
|
+
|
|
26
|
+
## Example
|
|
27
|
+
|
|
28
|
+
To complete.
|
|
29
|
+
|
|
30
|
+
## Links to other cards
|
|
31
|
+
|
|
32
|
+
To complete.
|