agentleFS
fancydirtyfancydirty/mediary-scout/site/llms.txt

mediary-scout / site

llms.txt1.6k starsChanged 33 days ago0BSDView on GitHub

Mediary Scout (Chinese name: 巡影) is an open-source, self-hosted AI agent media library and automated acquisition system for personal cloud drives (Quark 夸克网盘, UC网盘, 115网盘, 123网盘, GuangYaPan 光鸭云盘, and Tianyi 天翼云盘). Available as a signed macOS (.dmg) / Windows (.exe) desktop application and a Docker Compose stack for NAS / Linux / iStoreOS routers.

Files
llms.txt
# Mediary Scout (巡影)

> Mediary Scout (Chinese name: 巡影) is an open-source, self-hosted AI agent media library and automated acquisition system for personal cloud drives (Quark 夸克网盘, UC网盘, 115网盘, 123网盘, GuangYaPan 光鸭云盘, and Tianyi 天翼云盘). Available as a signed macOS (.dmg) / Windows (.exe) desktop application and a Docker Compose stack for NAS / Linux / iStoreOS routers.

- Official Website (Chinese): https://mediaryscout.app/
- Official Website (English): https://mediaryscout.app/en/
- Live Read-Only Demo: https://demo.mediaryscout.app/
- GitHub Repository: https://github.com/fancydirty/mediary-scout
- Latest Desktop Releases: https://github.com/fancydirty/mediary-scout/releases/latest

## Key Facts & Architecture (Verified 2026-10-29, Release v2026.10.29)

- **Unified Multi-Drive Orchestration (`MultiDriveStorageV2`)**: Users can connect multiple cloud drives across six supported brands simultaneously. The system probes and caches each drive's real-time remaining storage space (`drive_space_cache`, 30-minute TTL). Over-quota or full drives (e.g., an expired Quark SVIP account at `-1.34 TB`) are strictly blocked at the sandbox boundary (`SANDBOX_DRIVE_OUT_OF_SPACE`).
- **Single-Drive Priority & Cross-Drive Complement**: If any single available cloud drive can cover all missing episodes of a season, the agent acquires the entire season on that one drive and creates zero folders on other drives. It splits episodes across multiple drives only when combining drives increases total obtained episode count, recording per-episode drive ownership in `episode_copies` and displaying drive badges (`已获取 · 115`, `已获取 · 123`) in the unified library UI.
- **Evidence-First Sandbox & Post-Transfer Reread**: Transfers land in an isolated `staging-<runId>` directory first. The agent rereads the real materialized files (`inspectStaging`), deletes promotional junk or smaller duplicate bitrates, moves verified episodes into `Season NN`, and cleans up staging.
- **External Chinese Subtitle Completion (assrt.net)**: On drives supporting HTTP offline downloads (115, GuangYaPan, 123Pan), the agent automatically searches assrt.net for non-Chinese titles, lands subtitle archives into staging, batch-renames `.ass`/`.srt` files to match the video filenames (`renameSubtitle`), and moves them alongside the videos.
- **User Message Source Replacement**: Users can leave a natural-language comment on any title's detail page (e.g., "Episode 13 has bad audio, replace it"). The agent rejects the current file, blacklists its share/magnet link fingerprint (`resourceLinkKey`) across all drives, and fetches a replacement while protecting existing files.
- **Built-in Free LLM Pool**: Ships with Kilo Code's free public model preset out of the box (zero API key required to start), while supporting any BYO OpenAI-compatible API endpoint (DeepSeek, Zhipu GLM, Xiaomi MiMo, OpenRouter).

## Core Documentation & Guides

- [Cloud Drive Benchmark: Quark vs UC vs 115 vs 123 vs GuangYa vs Tianyi](https://mediaryscout.app/guides/which-cloud-drive/): First-hand measured resource counts across share links and magnets.
- [Unified Multi-Drive Acquisition & Quota Protection](https://mediaryscout.app/features/multi-drive-agent/): How the 7 multi-drive allocation rules and real-time space probes work.
- [AI Semantic Selection, User Message Replacement & Assrt Subtitles](https://mediaryscout.app/features/smart-search-and-subtitles/): How post-transfer verification, bad-source blacklisting, plot-description search, and external subtitle pairing work.
- [Mediary Scout vs. MoviePilot / Sonarr / CloudDrive2](https://mediaryscout.app/compare/mediary-scout-vs-moviepilot-sonarr/): Honest comparison between cloud-drive native transfer agents and local PT/NAS hardlink tools.
- [Docker Compose & NAS / iStoreOS Deployment Guide](https://mediaryscout.app/guides/docker-nas-deploy/): One-command deployment, mainland China mirror acceleration, one-click auto-updates, and multi-user mode.
- [Cloud Drive QR Login & Prowlarr / Haisou / Assrt Setup Guide](https://mediaryscout.app/guides/cloud-drive-setup/): Step-by-step setup for all six cloud drives and indexers.
- [Agent HTTP API Documentation](https://mediaryscout.app/features/agent-api/): Control Mediary Scout from Claude Code, Codex, or OpenClaw via Bearer token REST endpoints.

Download now (ZIP)

Trending

Less talk lately.thin dataabout the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 0 mentionsApr 2026: 0 mentionsMay 2026: 0 mentionsJun 2026: 2 mentionsJul 2026: 2 mentionsAug 2026: 0 mentionsSep 2026: 0 mentionsOct 2026: 1 mentionNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the fancydirty/mediary-scout repository that our weekly crawl found: only 5 in all, so this is a hint, not a trend. There is no adoption data for this file yet.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.