Cited Recall

by CaseLedgerself serveSetup: 1-2 hours0 downloads

Cited Recall turns scattered notes into something answerable. First run maps where your knowledge lives - Project files, a Drive folder, Notion pages, an Obsidian vault, a screenshots folder - then runs a guided ingestion pass: inventory, scope triage, retrievable renaming, splitting oversized documents, and a manifest with dates and authority order. After that you ask questions in plain English and every claim cites its file. When the corpus lacks the answer you get not-in-your-notes, not a guess. You configure sources, exclusions, answer style and privacy mode.

Included assets

second-brain.skill - installable Claude skill package
setup-guide.md - installation walkthrough for Claude app Projects and Claude Code
flow.md - end-to-end flow diagram and step reference
qa-checklist.md - validation checklist before first production use
useful-commands.md - command phrases for setup, daily use, and troubleshooting
research-notes.md - cited research behind the ROI benchmark

Capabilities

knowledge-retrievaldocument-qanote-synthesis

Setup requirements

Cowork or Claude Code - both persist your settings between sessions; the Claude app runs the skill but cannot save your setup
Code execution and file creation enabled in Claude settings - the skill installs via Settings > Customize > Skills
At least one knowledge source with a concrete pointer - a Project knowledge upload, a named Drive folder, specific Notion pages, a vault path, or a screenshots folder
Claude Code specifically if a local Obsidian vault or local folder must be read without uploading
Google Drive or Notion connector enabled if the corpus lives in those tools
A decision on scope and exclusions before ingestion - credentials, financial identifiers, and other people's confidential material must stay out of the corpus
Time for the first ingestion pass, including screenshot transcription for image-heavy corpora

Support includes

Self-serve setup guide covering Cowork, the Claude app, and Claude Code
Ingestion playbook for naming, splitting, screenshot transcription, exclusions, and capacity limits
Retrieval-failure troubleshooting ladder for when a known fact is not being found
Useful command reference for setup, daily use, and corpus maintenance

Implementation context

Target use case

AI personal knowledge base question answering

educationSave an estimated 20-36 minutes per working day on searching for and re-creating knowledge you already have

Package contents

v1.0.0

second-brain.skill
other · 31 KB
installable Claude skill package
Setup Guide
document · 10 KB
installation walkthrough for Claude app Projects and Claude Code
Flow
document · 9 KB
end-to-end flow diagram and step reference
QA Checklist
document · 4 KB
validation checklist before first production use
Useful Commands
document · 3 KB
command phrases for setup, daily use, and troubleshooting
Research Notes
document · 7 KB
cited research behind the ROI benchmark

Cited Recall

by CaseLedgerself serveSetup: 1-2 hours0 downloads

Included assets

second-brain.skill - installable Claude skill package

setup-guide.md - installation walkthrough for Claude app Projects and Claude Code

flow.md - end-to-end flow diagram and step reference

qa-checklist.md - validation checklist before first production use

useful-commands.md - command phrases for setup, daily use, and troubleshooting

research-notes.md - cited research behind the ROI benchmark

Cited Recall

Included assets

Capabilities

Setup requirements

Support includes

Implementation context

AI personal knowledge base question answering

Package contents

Browse automation packs that are ready to buy

Cited Recall

Included assets

Capabilities

Setup requirements

Support includes

Implementation context

AI personal knowledge base question answering

Package contents

Browse automation packs that are ready to buy