@charlie.act7/canvas-mcp-server 1.1.3 → 1.2.0

package/.mcp.json

@@ -1,10 +1,12 @@
 {
-  "canvas": {
-    "command": "npx",
-    "args": ["-y", "@charlie.act7/canvas-mcp-server"],
-    "env": {
-      "CANVAS_API_TOKEN": "${CANVAS_API_TOKEN}",
-      "CANVAS_API_DOMAIN": "${CANVAS_API_DOMAIN}"
+  "mcpServers": {
+    "canvas": {
+      "command": "npx",
+      "args": ["-y", "@charlie.act7/canvas-mcp-server"],
+      "env": {
+        "CANVAS_API_TOKEN": "${CANVAS_API_TOKEN}",
+        "CANVAS_API_DOMAIN": "${CANVAS_API_DOMAIN}"
+      }
     }
   }
 }

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Charlie Cárdenas Toledo
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/README.es.md

@@ -3,13 +3,26 @@
 [![npm version](https://img.shields.io/npm/v/@charlie.act7/canvas-mcp-server)](https://www.npmjs.com/package/@charlie.act7/canvas-mcp-server)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
-**¡Lleva tu aula virtual de Canvas al siguiente nivel con Inteligencia Artificial!** 🚀
+¡Lleva tu aula virtual de Canvas al siguiente nivel con Inteligencia Artificial! 🚀
 
-Este proyecto es un servidor de **Model Context Protocol (MCP)** para **Canvas LMS**. En palabras sencillas, funciona como un "traductor" o puente que permite a asistentes de Inteligencia Artificial (como Claude Desktop, Claude Code, Cursor, etc.) comprender y realizar acciones en tus cursos de Canvas mediante instrucciones en lenguaje natural.
+Este proyecto es un servidor de **Model Context Protocol (MCP)** para **Canvas LMS**. Funciona como un puente que permite a asistentes de Inteligencia Artificial (como Claude Desktop, Claude Code, Cursor, etc.) consultar y administrar tus cursos de Canvas mediante instrucciones en lenguaje natural.
 
 ---
 
-## 🔍 ¿Cómo funciona?
+## Tabla de Contenidos
+- [¿Cómo funciona?](#cómo-funciona)
+- [Ejemplos de Uso](#ejemplos-de-uso)
+- [Guía de Instalación](#guía-de-instalación)
+  - [Paso 1: Obtener credenciales de Canvas](#paso-1-obtener-credenciales-de-canvas)
+  - [Paso 2: Configurar tu Cliente de IA](#paso-2-configurar-tu-cliente-de-ia)
+- [Configuración por Consola (CLI)](#configuración-por-consola-cli)
+- [Herramientas y Recursos Soportados](#herramientas-y-recursos-soportados)
+- [Desarrollo Local](#desarrollo-local)
+- [Licencia](#licencia)
+
+---
+
+## ¿Cómo funciona?
 
 Cuando utilizas este servidor, la comunicación fluye de la siguiente manera:
 
@@ -23,40 +36,43 @@ graph LR
     AI -->|Respuesta amigable| User
 ```
 
-1. **Tú le pides algo a Claude** (por ejemplo: *"Crea una tarea para el próximo viernes"*).
-2. **Claude detecta la intención** y se comunica con el **Canvas MCP Server** enviándole los parámetros necesarios.
+1. **Tú le pides algo al asistente** (por ejemplo: *"Crea una tarea para el próximo viernes"*).
+2. **El asistente detecta la intención** y se comunica con el **Canvas MCP Server** enviándole los parámetros necesarios.
 3. **El servidor realiza la llamada segura** a la API de Canvas.
 4. **Canvas procesa la acción** y devuelve el resultado.
-5. **Claude te confirma el éxito de la operación** en lenguaje natural.
+5. **El asistente te confirma el éxito de la operación** en lenguaje natural.
 
 ---
 
-## 💡 ¿Qué puedes pedirle a tu Asistente?
+## Ejemplos de Uso
 
-¡Cualquier tarea administrativa o de consulta que harías manualmente en Canvas! Aquí tienes algunos ejemplos de lo que puedes pedirle:
+Aquí tienes algunos ejemplos de consultas y acciones reales que puedes pedirle a tu asistente:
 
-### 📖 Para Consultar Información
-* 💬 *"¿Qué cursos tengo activos este semestre?"*
-* 💬 *"Muéstrame las entregas pendientes de calificar para la Tarea 2 en Biología."*
-* 💬 *"¿Cuáles son los estudiantes del Grupo B de Química?"*
-* 💬 *"¿Qué cuestionarios o exámenes tenemos programados para esta semana?"*
+> [!TIP]
+> **Ahorro de Tokens y Eficiencia:** Siempre que sea posible, especifica el ID o la URL directa de Canvas (por ejemplo, `https://[tu_institucion].instructure.com/courses/[codigo_curso]/assignments/[codigo_actividad]`) en tus instrucciones. Esto evita que la IA tenga que buscar y escanear todos tus recursos, lo que resulta en respuestas mucho más rápidas y un ahorro significativo de tokens.
 
-### ✍️ Para Administrar y Crear Contenido
-* 💬 *"Crea una nueva tarea llamada 'Ensayo final de Historia' con fecha de entrega para el próximo viernes."*
-* 💬 *"Publica un anuncio en el curso de Matemáticas avisando que la clase de mañana será virtual."*
-* 💬 *"Crea un nuevo módulo llamado 'Semana 5: Introducción a React' en mi curso."*
+### 📖 Para Consultar Información y Auditar Cursos
+* 💬 *"¿Qué cursos tengo activos este semestre? Verifica si existen múltiples paralelos o secciones."*
+* 💬 *"Muéstrame las entregas pendientes de calificar para la actividad 'Ensayo 1: Introducción a la Sociología' en Sociología 101."*
+* 💬 *"¿Cuáles son los estudiantes registrados en el Grupo A de la clase de Química?"*
+* 💬 *"Verifica si la tarea 'Propuesta de Proyecto' tiene una rúbrica activa asociada. Si es así, obtén sus criterios."*
 
-### 💯 Para Calificar y Dar Retroalimentación
-* 💬 *"Califica el ensayo de María con un 90/100 y agrégale un comentario que diga: 'Excelente análisis crítico, continúa así'."*
-* 💬 *"Genera un reporte de las notas actuales de los alumnos del curso de Física."*
+### ✍️ Para Administrar y Crear Contenido Académico
+* 💬 *"Crea un nuevo módulo llamado 'Semana 1: Fundamentos' en mi curso."*
+* 💬 *"Agrega un Subencabezado 'LECTURAS OBLIGATORIAS' dentro del módulo 'Semana 1' y enlaza la página del sílabo."*
+* 💬 *"En mi curso de Negocios, crea una tarea llamada 'Estudio de Caso 1: Análisis de Mercado'. Agrega una tabla de instrucciones con columnas para Criterios, Requisitos y Puntaje."*
+
+### 💯 Gestión de Calificaciones y Asistencia
+* 💬 *"Para la tarea 'Estudio de Caso 1', busca a todos los estudiantes que no hayan entregado. Ponles un 0 de nota y agrégales el comentario: 'Actividad no entregada. Si tienes una justificación válida, por favor comunícate con el docente.'"*
+* 💬 *"Califica la entrega de Juan en 'Ensayo 1' con un 90 basado en la rúbrica, y agrega un comentario: '¡Excelente trabajo! El análisis está muy bien estructurado, aunque podrías profundizar un poco más en las conclusiones. ¡Sigue así!'"*
 
 ---
 
-## 🛠️ Guía de Instalación Paso a Paso
+## Guía de Instalación
 
 Para conectar tu asistente de IA a Canvas, necesitas configurar **dos cosas**: tus credenciales de Canvas y el cliente de IA (como Claude).
 
-### Paso 1: Obtener tus credenciales de Canvas LMS
+### Paso 1: Obtener credenciales de Canvas
 Para que el servidor pueda actuar en tu nombre, necesita permiso:
 1. Inicia sesión en tu cuenta de **Canvas LMS**.
 2. Dirígete a **Cuenta (Account)** ➡️ **Configuración (Settings)** en el menú lateral.
@@ -105,7 +121,7 @@ Luego, configura tus credenciales de forma interactiva:
 
 ---
 
-## 💻 Configuración Interactiva por Consola (CLI)
+## Configuración por Consola (CLI)
 Si prefieres configurar las credenciales de manera local e interactiva en tu terminal para desarrollo, puedes ejecutar:
 ```bash
 npx @charlie.act7/canvas-mcp-server config
@@ -114,8 +130,12 @@ Esto te pedirá el dominio y tu API token paso a paso, guardándolos de forma se
 
 ---
 
+## Herramientas y Recursos Soportados
+
 <details>
-<summary>🛠️ <b>Ver Lista Detallada de Herramientas Soportadas (Técnico)</b></summary>
+<summary><b>Ver Lista Detallada de Herramientas y Recursos Soportados (Técnico)</b></summary>
+
+### Lista de Herramientas
 
 El servidor expone internamente las siguientes herramientas organizadas por categorías:
 
@@ -144,7 +164,7 @@ Para clientes compatibles con recursos directos:
 
 ---
 
-## 🛠️ Desarrollo Local (Para Programadores)
+## Desarrollo Local
 
 Si deseas clonar este repositorio y hacer modificaciones:
 
@@ -169,5 +189,5 @@ Si deseas clonar este repositorio y hacer modificaciones:
 
 ---
 
-## 📄 Licencia
+## Licencia
 Este proyecto está bajo la licencia MIT. Creado por [Charlie Cárdenas Toledo](https://github.com/charlie-act7).

package/README.md

@@ -1,15 +1,30 @@
 # 🎓 Canvas LMS MCP Server
 
+[![CI](https://github.com/CharlieCardenasToledo/mcp-canvas-server/actions/workflows/ci.yml/badge.svg)](https://github.com/CharlieCardenasToledo/mcp-canvas-server/actions/workflows/ci.yml)
 [![npm version](https://img.shields.io/npm/v/@charlie.act7/canvas-mcp-server)](https://www.npmjs.com/package/@charlie.act7/canvas-mcp-server)
+[![npm downloads](https://img.shields.io/npm/dm/@charlie.act7/canvas-mcp-server)](https://www.npmjs.com/package/@charlie.act7/canvas-mcp-server)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
-**Bring AI to your Canvas Virtual Classroom!** 🚀
+Bring AI to your Canvas Virtual Classroom! 🚀
 
-This project is a **Model Context Protocol (MCP)** server for **Canvas LMS**. In simple terms, it acts as a "translator" or bridge that allows AI assistants (like Claude Desktop, Claude Code, Cursor, etc.) to understand and perform actions in your Canvas courses using natural language.
+This project is a **Model Context Protocol (MCP)** server for **Canvas LMS**. It acts as a bridge that allows AI assistants (like Claude Desktop, Claude Code, Cursor, etc.) to query and manage your Canvas courses using natural language.
 
 ---
 
-## 🔍 How It Works
+## Table of Contents
+- [How It Works](#how-it-works)
+- [Use Cases & Examples](#use-cases--examples)
+- [Setup Guide](#setup-guide)
+  - [Step 1: Obtain Canvas Credentials](#step-1-obtain-canvas-credentials)
+  - [Step 2: Connect to your AI Client](#step-2-connect-to-your-ai-client)
+- [CLI Configuration](#cli-configuration)
+- [Supported Tools & Resources](#supported-tools--resources)
+- [Local Development](#local-development)
+- [License](#license)
+
+---
+
+## How It Works
 
 When you interact with the server, communication flows as follows:
 
@@ -23,40 +38,64 @@ graph LR
     AI -->|Friendly Answer| User
 ```
 
-1. **You ask Claude for something** (e.g., *"Create an assignment due next Friday"*).
-2. **Claude detects your intent** and communicates with the **Canvas MCP Server**, sending the required parameters.
+1. **You ask the AI** (e.g., *"Create an assignment due next Friday"*).
+2. **The AI detects your intent** and communicates with the **Canvas MCP Server**, sending the required parameters.
 3. **The server makes a secure call** to the official Canvas API.
-4. **Canvas processes the action** and returns the result.
-5. **Claude confirms the success of the action** back to you in plain, natural language.
+4. **Canvas processes the action** and returns the response.
+5. **The AI confirms the success of the action** back to you in plain, natural language.
 
 ---
 
-## 💡 What Can You Ask Your Assistant?
+## Use Cases & Examples
+
+Here are some realistic, everyday prompts you can use with your AI assistant:
+
+> [!TIP]
+> **Token Saving & Efficiency:** Whenever possible, specify the Canvas ID or the direct Canvas URL (e.g., `https://[your_institution].instructure.com/courses/[course_id]/assignments/[assignment_id]`) in your prompts. This prevents the AI from scanning all your courses/resources, leading to faster responses and substantial token savings.
 
-Anything you would normally do manually in Canvas! Here are some everyday examples:
+### 📖 Course Auditing & Querying
+* 💬 *"What active courses do I have this semester? Check if there are multiple active sections/parallels."*
+* 💬 *"Show me all ungraded submissions for 'Essay 1: Introduction to Sociology' in Sociology 101."*
+* 💬 *"Who is in Student Group A for the Chemistry class?"*
+* 💬 *"Does the assignment 'Project Proposal' have an active rubric associated? If so, retrieve its criteria."*
+* 💬 *"Search for everything related to 'photosynthesis' across my Biology course — assignments, pages, and discussions."*
 
-### 📖 Querying Information
-* 💬 *"What active courses do I have this semester?"*
-* 💬 *"Show me the ungraded submissions for Assignment 2 in Biology."*
-* 💬 *"Who is in Group B of my Chemistry class?"*
-* 💬 *"What quizzes or exams are scheduled for this week?"*
+### ✍️ Creating & Organizing Course Content
+* 💬 *"Create a new module named 'Week 1: Foundations' in my course."*
+* 💬 *"Add a SubHeader 'REQUIRED READINGS' inside the 'Week 1' module, and link the syllabus page to it."*
+* 💬 *"In my Business course, create an assignment called 'Case Study 1: Market Analysis'. Add an instructions table with columns for Criteria, Requirements, and Points."*
+* 💬 *"Create a threaded discussion topic in my course titled 'Weekly Reflection' and pin it to the top."*
 
-### ✍️ Creating and Managing Content
-* 💬 *"Create a new assignment named 'History Term Paper' due next Friday."*
-* 💬 *"Post an announcement in my Math course letting everyone know tomorrow's lecture is online."*
-* 💬 *"Create a new module named 'Week 5: Introduction to React' in my course."*
+### 💯 Grading & Absence Management
+* 💬 *"For assignment 'Case Study 1', find all students who haven't submitted their work. Assign them a grade of 0 and add the comment: 'Activity not submitted. Please contact the instructor if you have a valid excuse.'"*
+* 💬 *"Grade John's submission for 'Essay 1' with a 90 based on the rubric, and add a comment: 'Great job! The analysis is well-structured, though you could expand more on the conclusion. Keep it up!'"*
 
-### 💯 Grading & Giving Feedback
-* 💬 *"Grade Maria's essay with a 90/100 and add a comment saying: 'Excellent critical analysis, keep it up!'"*
-* 💬 *"Give me a summary of current student grades in the Physics course."*
+### 📊 Student Engagement & Analytics
+* 💬 *"Show me the activity analytics for my Calculus course — how active have students been this week?"*
+* 💬 *"Which students haven't been active in course 12345 in the last few days? I want to reach out to them."*
+* 💬 *"Get the analytics for student [ID] in my Biology course — how many page views and participations do they have?"*
+
+### 💬 Messaging & Communication
+* 💬 *"Send a private message to student [ID] reminding them their 'Project Proposal' is due tomorrow."*
+* 💬 *"How many unread messages do I have in my Canvas inbox?"*
+* 💬 *"Show me my last 10 inbox conversations."*
+
+### 👥 Enrollment & Student Management
+* 💬 *"Who is enrolled in my course? Show me students and TAs separately."*
+* 💬 *"Search for a student named 'Maria Gonzalez' in account 1."*
+* 💬 *"Enroll user [ID] as a TA in my Physics course."*
+
+### 🔄 Peer Reviews
+* 💬 *"List all peer review assignments for 'Research Paper' in course 12345."*
+* 💬 *"Manually assign student [ID] to review [other student ID]'s submission for 'Essay 2'."*
 
 ---
 
-## 🛠️ Step-by-Step Setup Guide
+## Setup Guide
 
 To connect your AI assistant to Canvas, you need to configure **two things**: your Canvas credentials and your AI client (like Claude).
 
-### Step 1: Obtain your Canvas credentials
+### Step 1: Obtain Canvas Credentials
 To let the server act on your behalf, it needs permission:
 1. Log in to your **Canvas LMS** account.
 2. Go to **Account** ➡️ **Settings** in the sidebar menu.
@@ -105,7 +144,7 @@ Then, configure your credentials interactively:
 
 ---
 
-## 💻 Interactive Console Configuration (CLI)
+## CLI Configuration
 If you prefer to configure your credentials locally for development, run:
 ```bash
 npx @charlie.act7/canvas-mcp-server config
@@ -114,27 +153,35 @@ This will guide you step-by-step to input your domain and API token, storing the
 
 ---
 
+## Supported Tools & Resources
+
 <details>
-<summary>🛠️ <b>View Detailed List of Supported Tools (Technical)</b></summary>
-
-The server exposes the following tools organized by category:
-
-| Category | Included Tools |
-|---|---|
-| **Courses** | List courses, get course details, set basic configuration |
-| **Modules** | List and manage course modules |
-| **Pages** | List pages, read page content HTML |
-| **Files** | List files uploaded to a course |
-| **Announcements** | List and create announcements |
-| **Assignments** | List, create, and update assignments; bulk update due dates |
-| **Submissions** | View student submissions and attachments |
-| **Grading** | Grade submissions, audit course grades |
-| **Quizzes** | List quizzes, manage questions, update quiz dates |
-| **Students** | Course roster, progress tracking, and student details |
-| **Groups** | List and manage student groups |
-| **Calendar** | List and create calendar events/reminders |
-| **Rubrics** | Create and manage grading rubrics |
-| **Communication** | Send direct messages, manage discussions and threads |
+<summary><b>View all 111 tools organized by category</b></summary>
+
+### Tool Inventory
+
+| Category | Tools | Description |
+|---|---|---|
+| **Courses** | `list_courses` · `create_course` · `update_course` · `get_syllabus` | Manage and configure courses |
+| **Modules** | `list_modules` · `create_module` · `update_module` · `delete_module` · `create_module_item` · `update_module_item` · `delete_module_item` | Full CRUD for modules and their items |
+| **Pages** | `list_pages` · `get_page_content` · `create_page` · `update_page` · `delete_page` | Manage wiki pages |
+| **Files & Folders** | `list_files` · `upload_file` · `update_file` · `delete_file` · `list_folders` · `create_folder` · `update_folder` · `delete_folder` | File management with folder support |
+| **Assignments** | `get_assignments` · `get_assignment` · `create_assignment` · `update_assignment` · `delete_assignment` · `update_assignment_dates` · `bulk_update_due_dates` · `list_assignment_groups` | Full assignment lifecycle |
+| **Submissions** | `get_submissions` · `get_submission` · `get_submission_comments` · `delete_submission_comment` · `submit_assignment` | View and manage student submissions |
+| **Grading** | `grade_submission` · `grade_multiple_submissions` · `audit_course` | Grade individually or in bulk |
+| **Rubrics** | `list_rubrics` · `get_rubric` · `create_rubric` · `update_rubric` · `create_rubric_association` | Build and attach grading rubrics |
+| **Quizzes (Classic)** | `list_quizzes` · `get_quiz` · `create_quiz` · `update_quiz` · `update_quiz_dates` · `list_quiz_questions` · `get_quiz_question` · `create_quiz_question` · `update_quiz_question` · `delete_quiz_question` · `create_quiz_group` | Classic Canvas quiz engine |
+| **New Quizzes (LTI)** | `create_new_quiz` · `update_new_quiz` · `delete_new_quiz` · `list_new_quiz_items` · `get_new_quiz_item` · `create_new_quiz_item` · `update_new_quiz_item` · `delete_new_quiz_item` | Modern LTI quiz engine (`/api/quiz/v1`) |
+| **Students** | `list_students` · `list_students_with_grades` · `get_student_grades` · `get_student_assignments` · `list_assignment_due_dates` | Roster and progress tracking |
+| **Enrollments** | `list_course_enrollments` · `enroll_user` · `remove_enrollment` · `get_user` · `get_profile` · `search_users` | Manage who is in your course |
+| **Groups** | `list_group_categories` · `create_group_category` · `list_groups_in_category` · `create_group` · `assign_unassigned_members` · `add_group_member` | Student group management |
+| **Discussions** | `list_discussions` · `get_discussion_entries` · `create_discussion` · `delete_discussion` · `post_discussion_reply` | Discussion boards |
+| **Announcements** | `list_announcements` · `post_announcement` · `update_announcement` | Course announcements |
+| **Conversations** | `list_conversations` · `get_conversation` · `get_conversation_unread_count` · `send_conversation` · `reply_to_conversation` | Private inbox messaging |
+| **Calendar** | `list_appointment_groups` · `get_appointment_group` · `create_appointment_group` · `update_appointment_group` · `delete_appointment_group` · `list_appointment_group_users` · `list_appointment_group_groups` · `get_next_appointment` | Scheduling and appointments |
+| **Analytics** | `get_course_analytics` · `get_student_analytics` · `get_course_activity_stream` · `search_course_content` | Engagement data and content search |
+| **Peer Reviews** | `list_peer_reviews` · `get_submission_peer_reviews` · `create_peer_review` · `delete_peer_review` | Configure and manage peer assessments |
+| **Health & Config** | `health_check` · `set_canvas_config` | Verify connection and update credentials at runtime |
 
 ### Supported MCP Resources
 For clients supporting direct resources:
@@ -144,7 +191,7 @@ For clients supporting direct resources:
 
 ---
 
-## 🛠️ Local Development (For Developers)
+## Local Development
 
 To clone this repository and modify the code:
 
@@ -169,5 +216,5 @@ To clone this repository and modify the code:
 
 ---
 
-## 📄 License
+## License
 This project is licensed under the MIT License. Created by [Charlie Cárdenas Toledo](https://github.com/charlie-act7).

package/dist/http-server.js

@@ -1,6 +1,10 @@
 import Fastify from "fastify";
 import swagger from "@fastify/swagger";
 import swaggerUi from "@fastify/swagger-ui";
+import * as fs from "node:fs/promises";
+import * as path from "node:path";
+import { AgentRunner } from "./services/agent-runner.js";
+import { GeminiRunner } from "./services/gemini-runner.js";
 function buildOpenApiSpec() {
     return {
         openapi: "3.1.0",
@@ -21,7 +25,12 @@ function buildOpenApiSpec() {
                     summary: "Health check",
                     responses: {
                         "200": {
-                            description: "Server is healthy"
+                            description: "Server is healthy",
+                            content: {
+                                "application/json": {
+                                    schema: { type: "object", properties: { ok: { type: "boolean" } } }
+                                }
+                            }
                         }
                     }
                 }
@@ -32,7 +41,20 @@ function buildOpenApiSpec() {
                     summary: "Privacy policy",
                     responses: {
                         "200": {
-                            description: "Privacy policy text"
+                            description: "Privacy policy text",
+                            content: {
+                                "application/json": {
+                                    schema: {
+                                        type: "object",
+                                        properties: {
+                                            service: { type: "string" },
+                                            effective_date: { type: "string" },
+                                            summary: { type: "array", items: { type: "string" } },
+                                            contact: { type: "string" }
+                                        }
+                                    }
+                                }
+                            }
                         }
                     }
                 }
@@ -43,7 +65,24 @@ function buildOpenApiSpec() {
                     summary: "List active Canvas courses",
                     responses: {
                         "200": {
-                            description: "Courses list"
+                            description: "Courses list",
+                            content: {
+                                "application/json": {
+                                    schema: {
+                                        type: "array",
+                                        items: {
+                                            type: "object",
+                                            properties: {
+                                                id: { type: "integer" },
+                                                name: { type: "string" },
+                                                course_code: { type: "string" },
+                                                workflow_state: { type: "string" },
+                                                enrollment_term_id: { type: "integer" }
+                                            }
+                                        }
+                                    }
+                                }
+                            }
                         }
                     }
                 }
@@ -86,7 +125,26 @@ function buildOpenApiSpec() {
                     ],
                     responses: {
                         "200": {
-                            description: "Assignments list"
+                            description: "Assignments list",
+                            content: {
+                                "application/json": {
+                                    schema: {
+                                        type: "array",
+                                        items: {
+                                            type: "object",
+                                            properties: {
+                                                id: { type: "integer" },
+                                                name: { type: "string" },
+                                                due_at: { type: "string", nullable: true },
+                                                unlock_at: { type: "string", nullable: true },
+                                                lock_at: { type: "string", nullable: true },
+                                                points_possible: { type: "number", nullable: true },
+                                                published: { type: "boolean", nullable: true }
+                                            }
+                                        }
+                                    }
+                                }
+                            }
                         }
                     }
                 }
@@ -527,7 +585,7 @@ function buildOpenApiSpec() {
         }
     };
 }
-export async function startHttpServer(client, host = "0.0.0.0", port = 3000) {
+export async function startHttpServer(client, host = "0.0.0.0", port = 3000, tools = []) {
     const app = Fastify({ logger: true });
     await app.register(swagger, {
         openapi: buildOpenApiSpec()
@@ -535,6 +593,11 @@ export async function startHttpServer(client, host = "0.0.0.0", port = 3000) {
     await app.register(swaggerUi, {
         routePrefix: "/docs"
     });
+    app.setErrorHandler((error, _request, reply) => {
+        const status = error.statusCode ?? 500;
+        const message = status >= 500 ? "Internal server error" : error.message;
+        reply.code(status).send({ error: message });
+    });
     app.get("/health", async () => ({ ok: true }));
     app.get("/privacy", async () => ({
         service: "Canvas MCP HTTP API",
@@ -756,5 +819,76 @@ export async function startHttpServer(client, host = "0.0.0.0", port = 3000) {
             results
         };
     });
+    // --- POST /chat ---
+    const ollamaRunner = new AgentRunner(client, tools, process.env.OLLAMA_HOST ?? "http://localhost:11434");
+    const GEMINI_API_KEY = process.env.GEMINI_API_KEY ?? "";
+    const geminiRunner = GEMINI_API_KEY
+        ? new GeminiRunner(GEMINI_API_KEY, client, tools)
+        : null;
+    app.post("/chat", async (request, reply) => {
+        const { message, model, mode, provider = "ollama", gemini_key } = request.body ?? {};
+        if (!message || typeof message !== "string" || !message.trim()) {
+            return reply.code(400).send({ error: "El campo 'message' es requerido." });
+        }
+        if (tools.length === 0) {
+            return reply.code(503).send({ error: "El servidor no tiene herramientas cargadas. Inicia con 'serve-http'." });
+        }
+        if (provider === "gemini") {
+            const apiKey = gemini_key || GEMINI_API_KEY;
+            if (!apiKey) {
+                return reply.code(400).send({ error: "Se necesita gemini_key en el body o GEMINI_API_KEY en el entorno." });
+            }
+            const runner = gemini_key
+                ? new GeminiRunner(gemini_key, client, tools)
+                : geminiRunner;
+            return runner.run(message.trim(), { model });
+        }
+        return ollamaRunner.run(message.trim(), { model, mode });
+    });
+    // --- POST /audit ---
+    const SKILL_PATH = process.env.AUDIT_SKILL_PATH ?? path.resolve(process.env.USERPROFILE ?? process.env.HOME ?? ".", "Proyectos Personales", "material-docente - UIDE", ".claude", "commands", "canvas-module-auditor-uide-v2.md");
+    app.post("/audit", async (request, reply) => {
+        const { course_id, week, course_name, provider = "gemini", model, gemini_key } = request.body ?? {};
+        if (!course_id || !week) {
+            return reply.code(400).send({ error: "course_id y week son requeridos." });
+        }
+        if (tools.length === 0) {
+            return reply.code(503).send({ error: "El servidor no tiene herramientas cargadas." });
+        }
+        // Leer el skill de auditoría
+        let skillContent;
+        try {
+            skillContent = await fs.readFile(SKILL_PATH, "utf-8");
+        }
+        catch {
+            return reply.code(500).send({
+                error: `No se encontró el archivo de skill en: ${SKILL_PATH}. Define AUDIT_SKILL_PATH como variable de entorno.`
+            });
+        }
+        // Construir el mensaje de auditoría
+        const courseLabel = course_name ? `${course_name} (course_id: ${course_id})` : `course_id: ${course_id}`;
+        const auditMessage = `Ejecuta una auditoría instruccional completa de la Semana ${week} del curso ${courseLabel}. ` +
+            `Sigue exactamente el flujo de los pasos 1 al 4 definidos en las instrucciones. ` +
+            `Genera el reporte markdown completo con todas las secciones y la calificación global.`;
+        if (provider === "gemini") {
+            const apiKey = gemini_key || GEMINI_API_KEY;
+            if (!apiKey) {
+                return reply.code(400).send({ error: "Se necesita gemini_key en el body o GEMINI_API_KEY en el entorno." });
+            }
+            const runner = gemini_key
+                ? new GeminiRunner(gemini_key, client, tools)
+                : geminiRunner;
+            const result = await runner.run(auditMessage, {
+                model: model ?? "gemini-2.5-flash",
+                systemPrompt: skillContent
+            });
+            return { ...result, course_id, week };
+        }
+        const result = await ollamaRunner.run(auditMessage, {
+            model,
+            systemPrompt: skillContent
+        });
+        return { ...result, course_id, week };
+    });
     await app.listen({ host, port });
 }