ydy0615/llm-in-text
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: update AGENTS.md with new components and API details; enhance FileContent.vue with Plyr video support; add VideoPlayer.vue for video playback; improve MilkdownEditor.vue for video file handling; refine TTSPlayer.vue volume controls; optimize settings store for localStorage; implement video upload support in uploadBlock.js; create utility functions for Plyr player management.

Browse Source
This commit is contained in:
“ydy0615”
2026-06-13 11:03:40 +08:00
parent 2283020e51
commit 4813196b0a
16 changed files with 767 additions and 275 deletions
Download Patch File Download Diff File Expand all files Collapse all files
CLAUDE.md
+6 -1
View File
@@ -2,15 +2,20 @@
This file provides guidance to Claude Code (claude.ai/code) when working with this repository.
**Project version: v0.2.0** (full-stack architecture upgrade since b82c6d3)
## Project Overview
**LLM in Text** is an AI-powered Markdown editor built with Vue 3 + Vite (frontend) and FastAPI + Python + Ollama (backend). It provides real-time AI completion suggestions, OCR image recognition, document conversion (PDF/DOCX/PPTX to Markdown), and TTS text-to-speech.
**LLM in Text** is an AI-powered Markdown editor built with Vue 3 + Vite (frontend) and FastAPI + Python + Ollama + Redis Streams (backend). It provides real-time AI completion suggestions, OCR image recognition, document conversion (PDF/DOCX/PPTX to Markdown), TTS text-to-speech, web search integration (via SearXNG + Firecrawl), CAPTCHA verification, and async job queue processing.
- Completion interface uses plain POST/JSON (not SSE). Frontend sends `X-Request-Id` and calls `/v1/completions/cancel` on abort.
- AI completion is disabled when the document exceeds 32 KB (enforced both in UI and plugin layer).
- OCR text and doc-block content are injected into completion context as hidden context — they should NOT be rendered as visible user text.
- `/v1/convert` supports txt, docx, pptx, pdf via MarkItDown. Non-txt files go through Markdown sanitization (image removal, newline compression).
- `/v1/export/pdf` is called from the frontend but may not be implemented on the backend — verify before debugging PDF export.
- **Task queue architecture (v0.2.0 new)**: Backend shifted from synchronous endpoints to Redis Streams async queue. `job_system.py` defines JOB_TYPES (completion/pro_completion/web_search/compress/ocr/convert/tts/asr), `worker.py` consumes the queue, `job_handlers.py` registers handlers per type. All tasks support concurrency control, rate limiting, and circuit breakers.
- **Session tracking**: `session_store.py` provides InMemorySessionStore (dev) and PostgresSessionStore (prod), tracking request identity via session_hash + ip_hash.
- **Risk control system**: `risk_config.py` + `risk_control.py` implement rate limiting (sliding window), concurrency limits, circuit breaker pattern, and budget tracking — all thresholds via environment variables.
## Quick Start