Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .claude/agents/notebook-executor.md
Original file line number Diff line number Diff line change
Expand Up @@ -565,7 +565,7 @@ Task(
prompt="""
Agent notebook-executor.

Notebook: MyIA.AI.Notebooks/SymbolicAI/Argument_Analysis/Argument_Analysis_Executor.ipynb
Notebook: MyIA.AI.Notebooks/SymbolicAI/Argument_Analysis/Argumentation-08b-Executor-Python.ipynb
Mode: full
Kernel: python3
Env vars: {"BATCH_MODE": "true"}
Expand Down
2 changes: 1 addition & 1 deletion MyIA.AI.Notebooks/Probas/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -637,7 +637,7 @@ Cette série ancre mathématiquement ses résultats phares dans un assistant de
| QC ↔ Probas | `kelly_lean` | Fraction risquée `f* = μ−σ²/2` sous log-bienveillance (`0 sorry`) | [`Kelly_companion.ipynb`](../QuantConnect/kelly_lean/Kelly_companion.ipynb) |
| GameTheory ↔ Probas | `game_theory_lean` (Arrow) | Impossibilité d'Arrow (5 axiomes ⇒ dictature) | [`01-Arrow-Impossibility-Theorem.ipynb`](../GameTheory/SocialChoice/01-Arrow-Impossibility-Theorem.ipynb) |
| Search ↔ Probas | `search_lean` | Consistance heuristique `h ≤ h*` ⇒ optimalité `A*` | hub Search (cf [`search_lean/`](../Search/search_lean/)) |
| SymbolicAI ↔ Probas | `argumentation_lean` | Extension Dung (`grounded`/`preferred`/`stable`) par cadre formel | [`Argument_Analysis_Dung_AF_Semantics.ipynb`](../SymbolicAI/Argument_Analysis/Argument_Analysis_Dung_AF_Semantics.ipynb) |
| SymbolicAI ↔ Probas | `argumentation_lean` | Extension Dung (`grounded`/`preferred`/`stable`) par cadre formel | [`Argumentation-03-Dung-AF-Semantics-Python.ipynb`](../SymbolicAI/Argument_Analysis/Argumentation-03-Dung-AF-Semantics-Python.ipynb) |

```mermaid
flowchart LR
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@
"# Agent InformalAnalysisAgent (définitions)\n",
"\n",
"\n",
"[Retour au README](../README.md) | [Précédent : Init](./Argument_Analysis_Agentic-0-init.ipynb) | [Suivant : PL Agent](./Argument_Analysis_Agentic-2-pl_agent.ipynb)\n",
"[Retour au README](../README.md) | [Précédent : Init](./Argumentation-00-Setup-Tweety-Python.ipynb) | [Suivant : PL Agent](./Argument_Analysis_Agentic-2-pl_agent.ipynb)\n",
"\n",
"Dans ce notebook, nous definissons l'InformalAnalysisAgent, un agent specialise dans l'analyse informelle du discours. Cet agent identifie les arguments principaux d'un texte et detecte les sophismes en explorant une taxonomie hiérarchique (CSV Argumentum) combinee a un LLM.\n",
"\n",
Expand All @@ -31,7 +31,7 @@
"* Prompt Sémantique (`prompt_identify_args_v7`) et Fonction Setup (`setup_informal_kernel`)\n",
"* Instructions Système (`INFORMAL_AGENT_INSTRUCTIONS` - V13)\n",
"\n",
"---\n",
"***\n",
"\n",
"### Contexte et Objectifs\n",
"\n",
Expand Down Expand Up @@ -1429,7 +1429,7 @@
"tags": []
},
"source": [
"---\n",
"***\n",
"\n",
"## Conclusion : InformalAnalysisAgent - Un Agent Hybride LLM + Knowledge Graph\n",
"\n",
Expand Down Expand Up @@ -1468,7 +1468,7 @@
"\n",
"### Utilisation dans le workflow global\n",
"\n",
"Cet agent sera instancié dans le notebook d'orchestration (`Argument_Analysis_Agentic-3-orchestration.ipynb`) et recevra des tâches du `ProjectManagerAgent`, typiquement:\n",
"Cet agent sera instancié dans le notebook d'orchestration (`Argumentation-07-Orchestration-Python.ipynb`) et recevra des tâches du `ProjectManagerAgent`, typiquement:\n",
"\n",
"```json\n",
"{\n",
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@
"# Agent PropositionalLogicAgent (définitions)\n",
"\n",
"\n",
"[Retour au README](../README.md) | [Précédent : Init](./Argument_Analysis_Agentic-0-init.ipynb) | [Suivant : Orchestration](./Argument_Analysis_Agentic-3-orchestration.ipynb)\n",
"[Retour au README](../README.md) | [Précédent : Init](./Argumentation-00-Setup-Tweety-Python.ipynb) | [Suivant : Orchestration](./Argumentation-07-Orchestration-Python.ipynb)\n",
"\n",
"Cet agent est specialise dans l'analyse logique formelle (logique propositionnelle) en utilisant la bibliotheque Java Tweety via JPype.\n",
"\n",
Expand Down Expand Up @@ -83,7 +83,7 @@
"| 6.3. Instructions Système | Règles strictes de comportement de l'agent |\n",
"| 6.4. Test (commente) | Cellule de test du plugin (optionnelle) |\n",
"\n",
"---"
"***"
]
},
{
Expand Down Expand Up @@ -170,7 +170,7 @@
"\n",
"> **Note technique** : Le callback est configure lors de `setup_pl_kernel()` en passant une reference a `StateManagerPlugin.log_query_result`.\n",
"\n",
"---"
"***"
]
},
{
Expand Down Expand Up @@ -531,7 +531,7 @@
"\n",
"> **Impact pedagogique** : Ce type de bug est courant lors de l'integration de composants. Toujours verifier les signatures de fonctions lors de callbacks!\n",
"\n",
"---"
"***"
]
},
{
Expand Down Expand Up @@ -685,7 +685,7 @@
"\n",
"> **Note pedagogique** : Les prompts V10 demandent explicitement d'eviter `>>` pour eviter les confusions. L'opérateur `=>` est plus standard en logique formelle.\n",
"\n",
"---"
"***"
]
},
{
Expand Down Expand Up @@ -1061,7 +1061,7 @@
"- Rendre les résultats formels **comprehensibles** pour un utilisateur non-expert\n",
"- Relier les deductions logiques au **sens original** des arguments\n",
"\n",
"---\n",
"***\n",
"\n",
"### Fonction `setup_pl_kernel()` - Configuration du Semantic Kernel\n",
"\n",
Expand Down Expand Up @@ -1109,7 +1109,7 @@
"\n",
"**Utilite pedagogique** : Permet d'auditer toutes les requêtes executees par l'agent pour analyse ou debugging.\n",
"\n",
"---"
"***"
]
},
{
Expand Down Expand Up @@ -1250,7 +1250,7 @@
"- Le **PM** commande (créé des tâches)\n",
"- Les **agents specialises** executent (et reportent via `add_answer`)\n",
"\n",
"---"
"***"
]
},
{
Expand Down Expand Up @@ -1508,7 +1508,7 @@
"\n",
"Ces instructions refletent une approche **defensive** : mieux vaut etre trop prescriptif que laisser l'agent improviser. Les LLMs sont puissants mais impredictibles. Les contraintes strictes garantissent la fiabilite du système multi-agents.\n",
"\n",
"---"
"***"
]
},
{
Expand Down Expand Up @@ -1825,7 +1825,7 @@
"\n",
"Pour une analyse d'arguments complete et automatisee.\n",
"\n",
"---\n",
"***\n",
"\n",
"**Note pedagogique finale:**\n",
"\n",
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,7 @@
"# Orchestration de la conversation multi-agents\n",
"\n",
"\n",
"[Retour au README](../README.md) | [Précédent : PL Agent](./Argument_Analysis_Agentic-2-pl_agent.ipynb) | [Suivant : Executor](./Argument_Analysis_Executor.ipynb)\n",
"[Retour au README](../README.md) | [Précédent : PL Agent](./Argument_Analysis_Agentic-2-pl_agent.ipynb) | [Suivant : Executor](./Argumentation-08b-Executor-Python.ipynb)\n",
"\n",
"Dans ce notebook, nous definissons les stratégies d'orchestration qui gouvernent la conversation collaborative entre les agents d'analyse rhetorique. Nous implementons `SimpleTerminationStrategy` (arret sur conclusion ou max tours) et `DelegatingSelectionStrategy` (designation explicite avec fallback vers le ProjectManagerAgent), ainsi que la fonction principale d'exécution avec isolation complete des instances.\n",
"\n",
Expand Down Expand Up @@ -140,11 +140,11 @@
"**Prérequis:**\n",
"\n",
"Ce notebook suppose que vous avez exécuté:\n",
"- `Argument_Analysis_Agentic-0-init.ipynb` (configuration état et StateManager)\n",
"- `Argumentation-00-Setup-Tweety-Python.ipynb` (configuration état et StateManager)\n",
"- `Argument_Analysis_Agentic-1-informal_agent.ipynb` (agent analyse informelle)\n",
"- `Argument_Analysis_Agentic-2-pl_agent.ipynb` (agent logique propositionnelle)\n",
"\n",
"> **Note**: Ce notebook ne s'exécute pas de manière autonome. Il est importé et utilisé par `Argument_Analysis_Executor.ipynb` qui orchestre l'ensemble du processus."
"> **Note**: Ce notebook ne s'exécute pas de manière autonome. Il est importé et utilisé par `Argumentation-08b-Executor-Python.ipynb` qui orchestre l'ensemble du processus."
]
},
{
Expand Down Expand Up @@ -1096,7 +1096,7 @@
"source": [
"### Interprétation du flux d'exécution\n",
"\n",
"Lorsque la cellule ci-dessus est exécutée (via le notebook `Argument_Analysis_Executor.ipynb`), plusieurs éléments permettent de suivre le déroulement de l'analyse:\n",
"Lorsque la cellule ci-dessus est exécutée (via le notebook `Argumentation-08b-Executor-Python.ipynb`), plusieurs éléments permettent de suivre le déroulement de l'analyse:\n",
"\n",
"**Informations affichées à chaque tour:**\n",
"\n",
Expand Down
Loading
Loading