About this skill

Problem

Personal notes, URLs, files, and short TODOs often scatter across chat, making later lookup weak on source traceability. Memex compiles conversational input into structured Wiki pages and reads index.md and log.md before writing to reduce duplicate pages and duplicate ingestion.

How It Works

Ingest: URLs, images/PDFs/documents, or longer text enter ingestion; short memos go to inbox.md.
Query: triggered by 问：, /q, or explicit questions, retrieving answers from the Wiki with traceable references.
Maintenance: /health-check, /maintain, /status, and /inbox support health checks, stats, and inbox review.
Constraints: raw/ is append-only; pages/ claims must trace to specific raw files; untraceable content is marked ⚠️ 推测, conflicts are marked ⚠️ CONFLICT, and missing answers are explicitly reported.

Boundaries

It expects a valid vault path and fits long-term accumulation of citable material. Do not force writes when sources are missing or initialization is incomplete; concurrent edits to index.md and log.md require protection.

Screenshots

Use Cases

Send product doc URLs and PDFs to compile sourced Wiki pages for later answers.
Use “Ask:” to query ingested technical notes and cite the raw source file.
Write short memos and TODOs to inbox, then review pending items with /inbox.
Run /health-check to inspect conflicts, speculation labels, and vault status.

Best For

Individual engineers maintaining technical docs: turning scattered URLs and attachments into traceable notes.
Research assistants organizing materials: compiling PDFs and long articles into sourced Wiki pages.
Knowledge workers managing TODOs: centralizing short memos in inbox for later review.
Developers maintaining LLM knowledge bases: checking conflicts, speculation labels, and vault status.

Related Skills

Huawei Cloud Global Search

@user_89241a1c/huaweicloud-global-search

Search Huawei Cloud official docs and product pages to find ECS, OBS, RDS, CCE product specs, parameters, documentation, and API references without login.

Knowledge Management 2026.08.29 ↓ 366

Multi-Ticket OCR Assistant

@user_e7bc6b23/ticket-ocr-skill

OCR-based recognition for movie, train, flight, and event tickets in images or PDFs, extracting key fields into Markdown or JSON reports.

Knowledge Management 2026.08.29 ↓ 369

Next From Knowledge

@user_15292d5a/yjkj-next-from-knowledge

Turns notes, research, and meeting summaries into actionable next moves, plans, decisions, experiments, and decision-changing gaps.

Knowledge Management 2026.08.30 ↓ 371

Wiki Creator Knowledge Base Manager

@user_5b28ea14/wiki-creator

A local wiki knowledge base manager that compiles raw documents into sourced, indexed Markdown pages with wikilinks, query support, and health checks.

Knowledge Management 2026.08.29 ↓ 314

Memex LLM Wiki Knowledge Base