Scan _now.md unprocessed sections and generate an HTML classification plan.
npx skills add https://github.com/texiwustion/s-capture-loop --skill capture-loop
Please help me install this Agent Skill. Skill: capture-loop Source: https://github.com/texiwustion/s-capture-loop/tree/main Command: npx skills add https://github.com/texiwustion/s-capture-loop --skill capture-loop
This Skill turns scattered “unprocessed” sections in your running capture buffer into an auditable, self-contained HTML plan by scanning, classifying, and cross-referencing project documentation before you move anything.
_now.md for unprocessed sections: detects missing or changed [processed:... hash:...] markers and selects only what needs attention.project_docs files/dirs to enrich classification context and reduce guesswork.In Claude Code, run /capture-loop to generate today’s classification plan HTML for manual review.
High-intent search queries and answers about installing and using this skill.
To classify unprocessed markdown sections in a daily capture buffer, you can scan your now-file for missing or changed hash markers and route the content into reference notes, inbox fleeting notes, and a review queue. This workflow cross-references project docs to enrich classification context.
The best way to generate an audit-ready classification plan is to produce a self-contained HTML file that includes routing reasons and inline wikilink checks. This plan is generated after scanning and classifying your buffer, ensuring human-audited movement of content.
Hash-based change detection works by scanning your configured now-file for specific processed markers and comparing their hash values. It selects only the sections with missing or changed hashes, ensuring that previously processed content is ignored while new or modified entries are flagged for classification.
Yes, you can use regex routing to automatically move fleeting notes into an inbox. The classification system applies routing regex to sort content into predefined buckets such as what_how references, inbox notes, review queues, and stay-in-place narrative.
No, this capture classification workflow does not support automated cron-driven execution. It requires a two-stage safety workflow where you scan, classify, and generate a plan first, then execute the routing only after explicit human confirmation.
You need to cross-reference project documentation before routing capture notes to enrich the classification context and reduce guesswork. By reading configured project_docs files and directories, the system ensures that commands and SOPs are accurately routed to the correct reference notes.
forgefabrikGenerate hierarchical repository codemaps with hash-based change detection.
robinslangeClassify inbox notes by intent, cluster by topic, and promote mature notes.
TymorIbrahimCreate and manage hookify rules with YAML configuration and regex matching.
okwindsAudit repository compliance and generate markdown reports with Python scripts.