Unifies the new LangGraph-driven course-plan/media flow with robust provider fallbacks, admin AI provider settings, editable book-style materials, and strict entity isolation across LMS/course-plan APIs. Adds admin-only entity membership management in the Entities UI so users can switch linked entities directly from the platform.
Made-with: Cursor
Builds the §24 product on top of the LangGraph runtime from §22:
Phase A (Sources / RAG)
- encoach.course.plan.source model (file | url | text)
- SourceIndexer extracts PDF (pypdf), DOCX (python-docx), HTML, plain
text and embeds chunks via the existing pgvector pipeline scoped to
plan_id, so resources.search only returns the plan's own corpus
- Endpoints: list/create/upload/reindex/delete + plan-scoped retrieval
Phase B (Deliverables)
- services.deliverables.compute_deliverables walks the plan, derives
{planned, generated, ready} per week from material + media state
- GET /api/ai/course-plan/<id>/deliverables drives the new wizard
preview step and the live progress strip on the detail page
Phase C (Multi-modal media)
- encoach.course.plan.media model + MediaService:
audio: AWS Polly (default) or ElevenLabs
image: OpenAI DALL-E 3, capped per plan via system parameter
video: local ffmpeg subprocess (image + audio -> MP4 1280x720)
- Three new agent tools (media.synthesize_audio / generate_image /
compose_video), wired into course_week_materials and a new
course_media_director agent
- Endpoints per material + week-level batch generator
Phase D (Assignments)
- encoach.course.plan.assignment supports mode='batch' (op.batch) or
mode='students' (res.users), with due_date + message + state
- REST endpoints to list / create / delete assignments
Phase E (Student view)
- /api/student/course-plans + /api/student/course-plans/<id>
enforce visibility via assignment.expand_user_ids()
- New /student/course-plans list + read-only drilldown rendering
audio/image/video tiles from /web/content/<attachment_id>
Cross-cutting
- encoach.ai.tool.category: + media (so the new tools register)
- encoach.embedding gains a plan_id filter for plan-scoped RAG
- Wizard adds Sources + Multimedia steps; AdminCoursePlanDetail
rewritten with DeliverablesStrip + SourcesCard + AssignmentsCard +
per-material MediaDrawer
- ~280 new EN + AR i18n keys (full RTL coverage)
- smoke_course_plan.py exercises every phase via odoo-bin shell;
last run: PASS A/B/D/E + DALL-E 3 image (753 KB), Polly audio
fails cleanly when AWS creds aren't configured (expected)
Documentation: §24 added to docs/PROJECT_SUMMARY.md with phase-by-phase
artefact list, endpoints, smoke test, ops notes, and gotchas.
Made-with: Cursor
Based on UTAS GE1 Course Outline structure (Reading/Writing 10hrs + Listening/Speaking 8hrs)
New Models:
- encoach.course.plan.deliverable: Explicit learning outcome tracking by week/skill
- encoach.course.plan.resource.dep: Resource dependencies (textbooks, videos, etc.)
- encoach.course.plan.assignment: Assign plans to classes/students with progress tracking
- encoach.course.plan.assignment.deliverable: Per-student deliverable completion status
Extended Models:
- course.plan.material: Added media fields (media_type, media_asset_url, media_asset_id,
media_generation_prompt, media_metadata_json) for rich content
- New material types: video_lesson, audio_recording, image_visual, interactive, assessment
AI Agent Tools (agent_tools.py):
- deliverables.detect: Parse course outlines (like GE1 PDF) and extract structured outcomes
- deliverables.fetch: Get deliverables for AI to reference when generating
- resources.fetch: Check available resources before generating content
- resources.save: Persist resource dependencies
- media.suggest_visuals: AI suggests images/diagrams for materials
- media.generate_image: Generate educational images (DALL-E integration ready)
- media.generate_audio: Generate TTS audio (ElevenLabs/Polly integration ready)
- assignment.*: Create assignments and track progress
Pipeline Enhancements (course_plan_pipeline.py):
- generate_deliverables_from_outline(): Parse PDF/text outlines into structured deliverables
- generate_week_materials_with_resources(): Resource-aware content generation
- suggest_media_for_material(): AI visual aid suggestions
- generate_media_for_material(): Actual image/audio generation
New AI Agents (agents_defaults.xml):
- deliverable_detector: Parses GE1-style outlines, extracts deliverables week-by-week
- media_generator: Creates images/audio for teaching materials
- Updated course_planner & course_week_materials with resource tools
REST APIs (course_plan.py):
POST /api/ai/course-plan/<id>/deliverables/detect - Parse outline
GET /api/ai/course-plan/<id>/deliverables - List deliverables
PUT /api/ai/course-plan/deliverables/<id> - Update status
GET /api/ai/course-plan/<id>/resources - List resources
POST /api/ai/course-plan/<id>/resources - Add resource
POST /api/ai/course-plan/materials/<id>/media/suggest - Get visual suggestions
POST /api/ai/course-plan/materials/<id>/media/generate - Generate image/audio
POST /api/ai/course-plan/<id>/assignments - Assign to class/student
GET /api/ai/course-plan/<id>/assignments - List assignments
GET /api/ai/course-plan/assignments/<id> - Get with progress
PUT /api/ai/course-plan/assignments/<id>/deliverables/<del_id> - Update status
Security: Added ir.model.access.csv entries for all new models
Made-with: Cursor
Addresses the QA notes on the end-to-end approval flow. Highlights:
Backend
- encoach_ai/generation_submit: create encoach.approval.request on submit so
submitted exams actually reach the approver queue (previously only the
workflow id was stored on the exam, leaving approvers with an empty inbox).
Initial exam status flips to pending_approval instead of draft.
- encoach_exam_template/approval_workflows:
* /api/approval-users filters out students (user_type='student',
op_student_id set, share=True) so the approver dropdown only lists staff.
* _ser_request enriches requests with target_name / target_status and the
current stage approver for UI badges.
* list_requests supports mine=1 / requester=1 so approvers only see queue
items awaiting their action.
* approve_request / reject_request now transition the underlying
encoach.exam.custom to published / rejected on final approval.
- encoach.exam.custom.status: add pending_approval and rejected states to
match the approval workflow transitions.
Frontend UX
- GenerationPage: rubric field shows "Auto-graded — no rubric" for Listening
and Reading (rubrics only apply to Writing / Speaking). Divider dropdowns
now carry explicit help text explaining None / Line / Space / Page Break
and where to write prompts. Selecting an official exam structure
auto-populates the required tasks/sections/parts, the delete button is
hidden for essential tasks, and submit is blocked if the user dropped
below the structure's required count.
- RubricsPage: "Add Criterion" is now a dropdown with IELTS-specific
presets (Task Achievement, Coherence & Cohesion, Fluency & Coherence, …)
keyed off the selected skill, plus a "Custom (blank row)" fallback.
- AdminCourses: added tooltips and helper copy clarifying Difficulty vs
CEFR Level in the Create Course dialog.
- CustomExamCreate: validate the whole form before Save/Publish, surface
backend error messages (413/504/401) instead of a generic toast, read the
exam id from the correct response field, and avoid marking Publish as
failed when only the optional Save-as-Template step errors.
- ResourceManager / TeacherLibrary: upload toast now shows a concrete
reason (HTTP 413 / 504 / 401) instead of a generic "Upload failed".
Config (504 / 413 / resource upload fixes)
- odoo.conf + odoo-docker.conf: raise limit_time_cpu to 600s,
limit_time_real to 900s, limit_request to 128 MB, and memory caps so
/api/exam/generation/submit and multipart resource uploads stop hitting
504 / 413 during QA.
- frontend/Dockerfile (nginx): add client_max_body_size 128m, bump
proxy_read_timeout / proxy_send_timeout to 900s, and disable request
buffering so large AI/resource payloads stream through the proxy.
Made-with: Cursor
Frontend
- i18n: install tailwindcss-rtl, Cairo font, RTL-aware direction in index.css.
- Language toggle: localize aria-label / menu label, persist choice, update
document dir synchronously.
- Sidebar: add `side` prop so the drawer pins to the right in RTL; wire up
AdminLmsLayout, RoleLayout (student/teacher) and AppSidebar to pass
side = i18n.dir() === 'rtl' ? 'right' : 'left'.
- AdminLmsLayout: convert every nav item from hard-coded title to titleKey,
translate group labels (incl. the collapsible Training), breadcrumbs,
user menu (Profile / Settings / Logout), help button and toggle aria
labels; replace physical mr-/right- utilities with logical me-/end-.
- AI components (AiTipBanner, AiInsightsPanel, AiAlertBanner, AiSearchBar,
AiAssistantDrawer): apply dir="auto" at the container level, localize
titles, loading / error / empty states.
- Dashboards (admin / student / teacher): wrap numeric values in <bdi>,
localize dates via ar-EG, fix flex direction for KPI and assignment cards.
- UI primitives (breadcrumb, calendar, carousel, dropdown-menu, menubar,
context-menu, pagination, sidebar): flip chevrons in RTL via a scoped
CSS rule, swap pl-/pr-/ml-/mr- for ps-/pe-/ms-/me-.
- Add logical-direction helpers and bidirectional isolation classes.
Locales
- Expand en.ts and ar.ts with full `nav`, `sidebarGroup`, `breadcrumb`,
`userMenu`, `chrome`, `ai`, and dashboard key sets; keep key parity.
API client
- `api-client.ts` reads the active language from localStorage/i18n and sends
`Accept-Language` on every request so the backend can localize AI output.
Backend (encoach_ai)
- openai_service: add _LANGUAGE_NAMES, normalize_language, language-aware
system prompt injection for every OpenAI call.
- coach_service + controllers (coach_controller, ai_controller): thread
the requested language from headers / user locale down to OpenAIService.
- ai_feedback: fix latent registry error by pointing course_id at op.course
instead of the non-existent encoach.course.
Other
- .gitignore: ignore runtime odoo logs and local caches.
Made-with: Cursor
- Add per-section instructions field to exercise wizard with sensible defaults
- Group exercises by type with section headers showing instructions
- Pass type_instructions to backend AI prompt for context-aware generation
- Add instructions editing in exercise edit dialog
- Update Exercise interface to include instructions field
Made-with: Cursor
- Fix ELAI video generation (correct render endpoint, script splitting for 60s limit)
- Fix speaking script generation error handling and empty response display
- Add custom exam list API (GET /api/exam/custom/list)
- Add assignments REST API (list, create, get)
- Add rubrics REST API (list, create)
- Enhance Generation page: dynamic exam structures, auto-module selection, preview dialog, audio player
- Improve submit feedback with exam ID and status in toast notifications
- Fix ExamsListPage to show both custom exams and exam sessions
- Connect RubricsPage to backend API with fallback data
- Add Dockerfile, docker-compose.yml, requirements.txt for deployment
- Fix placement, grading, scoring, and auth controllers
- Add ErrorBoundary component for frontend resilience
- Add QA report and credentials documentation
Made-with: Cursor
