karpathy-wiki-capture

Standardize capture authoring for Karpathy Wiki integrations with canonical frontmatter.

98|24|Updated Apr 4, 2026
One-click install
npx skills add https://github.com/toolboxmd/karpathy-wiki --skill karpathy-wiki-capture
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: karpathy-wiki-capture
Source: https://github.com/toolboxmd/karpathy-wiki/tree/main/skills/karpathy-wiki-capture
Command: npx skills add https://github.com/toolboxmd/karpathy-wiki --skill karpathy-wiki-capture

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Establishes a canonical capture-authoring protocol for the Karpathy Wiki integration, detailing how to format a capture, how to invoke bin/wiki capture, body-size floors, and the subagent-report workflow. It ensures consistent provenance and traceability between wiki data sources and their derived captures.

Core Features & Use Cases

  • Defines the canonical frontmatter and required fields (title, capture_kind, evidence, body) and a consistent path into the wiki ingestion pipeline.
  • Describes how to route captures to the appropriate wiki (project vs main), how to trigger ingestion, and how to handle inbox-driven or drift-initiated ingestions.
  • Documents how subagent reports are ingested as captures when produced by research components.

Quick Start

Invoke bin/wiki capture with a title and kind, pipe the capture body to stdin, and allow the ingester to manage page creation or updating.

Frequently Asked Questions about karpathy-wiki-capture

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

FAQPage Schema
How do I standardize wiki capture authoring with canonical frontmatter?

To standardize wiki capture authoring, define a canonical frontmatter schema including title, capture_kind, evidence, and body fields. This enforces consistent provenance and traceability between wiki data sources and their derived captures.

How do I route wiki captures to the correct project or main wiki?

Routing wiki captures involves scoping capture kinds and using defined ingestion workflows to direct pages to the correct project or main wiki. This ensures deterministic writing and proper provenance tracking for each entry.

What is the workflow for ingesting subagent reports as wiki captures?

Subagent reports are ingested as captures when produced by research components. The workflow involves piping the capture body to stdin via the wiki capture command, allowing the ingester to manage page creation or updating automatically.

Does wiki ingestion enforce required fields and body-size floors for captures?

Yes, wiki ingestion enforces required frontmatter fields such as title, capture_kind, evidence, and body, alongside specific body-size floors. It also provides guidance for handling drift or inbox-based ingestion logistics.

What is the best way to handle inbox-driven or drift-initiated wiki ingestion?

Handling inbox-driven or drift-initiated ingestion requires following the protocol's trigger-based logistics. The system provides specific guidance for managing these edge cases to maintain deterministic writing and consistent provenance.