corpus-capture

Capture ad-hoc notes into a Lorekit corpus workbench inbox as timestamped Markdown fragments.

5|1|Updated Apr 15, 2026
One-click install
npx skills add https://github.com/GYF0311/lorekit --skill corpus-capture
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: corpus-capture
Source: https://github.com/GYF0311/lorekit/tree/main/skills/corpus-capture
Command: npx skills add https://github.com/GYF0311/lorekit --skill corpus-capture

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It solves the problem of losing or scattering quick notes (ideas, decisions, observations, temporary context) across projects by providing a reliable way to funnel them into a central Lorekit corpus workbench inbox without compiling long-term knowledge.

Core Features & Use Cases

  • Cross-project inbox capture: Stores fragments into the configured corpus workbench collection so multiple repositories can feed one knowledge base.
  • Append-only Markdown fragments: Creates a timestamped, append-only note under the inbox directory to preserve provenance and reduce accidental overwrites.
  • Safety-first workflow: Avoids mutating 知识库/, avoids running compile/sync steps, and uses frontmatter to keep capture metadata structured.
  • Use case: After a coding decision or an investigation while working on an issue, capture the rationale and key observations so you can later reconcile and synthesize them into long-term wiki knowledge.

Quick Start

Tell your AI to capture your current observation into the Lorekit corpus workbench inbox using the global-corpus.json configuration.

Frequently Asked Questions about corpus-capture

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I capture cross-project knowledge into a centralized inbox without compiling long-term wiki files?

Cross-project knowledge capture works by writing ad-hoc notes as append-only Markdown fragments into a centralized corpus workbench inbox, preventing unresolved ideas from being compiled into long-term wiki files.

What is the best way to preserve provenance for temporary observations across multiple repositories?

Preserving provenance for temporary observations requires writing timestamped, append-only Markdown fragments with structured inbox frontmatter into a shared corpus workbench directory, ensuring per-project tracking without accidental overwrites.

How do I record coding decisions and temporary context for later reconciliation?

Recording coding decisions involves funneling temporary context and observations into a centralized Lorekit corpus inbox, keeping them as uncompiled fragments until you are ready to synthesize them into wiki knowledge.

Does capturing inbox notes modify existing knowledge base directories or trigger sync operations?

Capturing inbox notes does not modify existing knowledge base directories or run compile and sync steps, ensuring a safety-first workflow that strictly isolates workbench-only capture from long-term knowledge compilation.

Do I need a global configuration file to capture notes into the Lorekit corpus workbench?

Yes, capturing notes into the Lorekit corpus workbench requires reading the global-corpus.json configuration file to determine the correct inbox path for writing the timestamped Markdown fragments.

When should I not use an append-only Markdown fragment approach for knowledge capture?

You should not use append-only Markdown fragments for finalized context, as this approach is designed exclusively for unresolved ideas and temporary context that should not be file-backed into long-term wiki knowledge.