knowledge-capture

Capture anything into structured Markdown notes and training-ready JSONL.

💾 Source code and 📓 TeX for AL Hayes, M Das, P Odom, and S Natarajan, "User Friendly Automatic Construction of Background Knowledge: Mode Construction from ER Diagrams." Knowledge Capture Conference 2017

Local raw capture layer for agent sessions, preserving decisions and handoffs as wiki/LLM knowledge-base ingestion.

Thinkly Chrome extension release distribution repository

End-of-session handoff that captures all knowledge, updates documentation, and prepares next-session prompts

Turns your long Claude sessions into a clean, structured "handoff note" so any new chat picks up exactly where you left off — with full context, decision history, active threads, preferences, setups, and approved or rejected paths already preserved. For the initiated, think of it as the smarter sibling to /compact.

Record yourself doing a task once, answer a few questions, and get a SKILL.md your AI agent can repeat. Local-first macOS recorder with an optional server; works with any OpenAI-compatible model.

Capture architecture from conversation. One pipeline, seven agent skills, a plain-language map anyone can read.

Clean video links into readable knowledge: transcript cleanup, summary, and Reality Check for Xiaohongshu, Douyin, and YouTube

Codex skill for capturing behavior-changing conversation steps into reusable Markdown notes.

A working proof-of-concept in knowledge-preserving AI-assisted development.

Arch-based live environment for AI-assisted work, with project-terminal scaffolding and local workflow logging.

Social-agent skill generator that turns guided interviews into reusable expression-agent skills for explaining ideas, judgments, and expertise.

I was obselete I couldn't hear the beat Staggerin about on me old mans feet I had one foot in the grave But now I'm nu rave! But now I'm nu rave!

Watch an operational process once, get editable process maps and runbooks. The capture layer for tribal knowledge.

Open-source Python toolkit for capturing expert decision logic as structured, reviewable knowledge cards.

Offline, consent-first capture of expert know-how. Transcribes a technician’s narrated task, structures it into verified steps with a local LLM, and lets juniors ask in plain language, abstaining when no expert has covered it. Stage I prototype for the ASUS UGen AI League 2026 (Workplace AI).

MindCache FYI browser extension — capture AI chat context from the web

Plantilla para relevar y documentar cómo trabaja una empresa con un agente de IA, bajo control documental IRAM-ISO 9001