reduce
Extract structured knowledge from source material. Comprehensive extraction is the default — every insight that serves the domain gets extracted. For domain-relevant sources, skip rate must be below 10%. Zero extraction from a domain-relevant source is a BUG. Triggers on "/reduce", "/reduce [file]", "extract insights", "mine this", "process this".
Profile →npx skills add majiayu000/claude-skill-registry --skill reduce --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Extract composable {vocabulary.note_plural} from source material into {vocabulary.notes}/.
How it works
- It defines a runtime configuration layer and a mission to perform comprehensive extraction, prioritizing core domain assertions, evidence, patterns, tensions, and enrichments.
- It enforces comprehensive extraction for domain-relevant sources and treats off-topic content via a selectivity gate.
- It outlines an execution pipeline: read source, check size (chunk if large), hunt for insights across defined extraction categories, classify candidates as OPEN or CLOSED, and output an extraction report. It provides a tiered search strategy:
- Tier 1: use
mcp__qmd__vector_searchwith query "[claim as sentence]" on collection{vocabulary.notes_collection}, limit 5 - Tier 2:
qmd vsearch "[claim as sentence]" --collection {vocabulary.notes_collection} -n 5 - Tier 3: keyword grep as a fallback
- Tier 1: use
- It requires user approval before creating files and supports a --handoff mode to create per-claim task files and update a queue, emitting a RALPH HANDOFF block.
- It includes explicit steps for source reading, large source handling, candidate classification, and extraction reporting.
When to use it
- Use when you need to extract comprehensive, domain-relevant {vocabulary.note_plural} from source material, guided by the extraction categories and the mission to externalize reasoning.
- Trigger phrases include "/reduce", "/reduce [file]", "extract insights", "mine this", "process this" as per the skill description.
What it can touch
- Tools listed: Read, Write, Grep, Glob, mcp__qmd__vector_search
- It references the collection
{vocabulary.notes_collection}for vector search, and uses commands likemcp__qmd__vector_search,qmd vsearch.
Caveats
- Implements a strict size handling rule: if source exceeds 2500 lines, it stops and plans chunks of 350-1200 lines.
- Zero extraction from a domain-relevant source is a BUG, per the mission constraints.
- The review relies on placeholders like {vocabulary.note_plural}, {vocabulary.domain}, and {vocabulary.notes_collection}, which must be resolved in the target environment.
- No outcomes are promised beyond the described extraction and reporting workflow; results depend on source content and user approval.
## Runtime Configuration (Step 0 — before any processing) Read these files to configure domain-specific behavior: 1. **`ops/derivation-manifest.md`** — vocabulary mapping, extraction categories, platform hints - Use `vocabulary.notes` for the notes folder name - Use `vocabulary.inbox` for the inbox folder name - Use `vocabulary.note` for the note type name in output - Use `vocabulary.note_plural` for the plural form - Use `vocabulary.reduce` for the process verb in output - Use `vocabulary.cmd_reflect` for the next-phase command name - Use `vocabulary.cmd_reweave` for the backward-pass command name - Use `vocabulary.cmd_verify` for the verification command name - Use `vocabulary.extraction_categories` for domain-specific extraction table - Use `vocabulary.topic_map` for MOC/topic map references - Use `vocabulary.topic_maps` for plural form 2. **`ops/config.yaml`** — processing depth, pipeline chaining, selectivity - `processing.depth`: deep | standard | quick - `processing.chaining`: manual | suggested | automatic - `processing.extraction.selectivity`: strict | moderate | permissive 3. **`ops/queue/queue.json`** — current task queue (for handoff mode) If these files don't exist (pr
- Runtime Configuration (Step 0 — before any processing)
- THE MISSION (READ THIS OR YOU WILL FAIL)
- The Core Distinction
- The Comprehensive Extraction Principle
- The Extraction Question (ask for EVERY candidate)
- INVALID Skip Reasons (these are BUGS)
- VALID Skip Reasons (rare)
- EXECUTE NOW
- Observation Capture (during work, not at end)
- Philosophy
- Extraction Categories
- What To Extract
- Category Detection Signals
- The Mission Lens (REQUIRED)
Get descriptions from existing notes
for f in {vocabulary.notes}/*.md; do
done
qmd vsearch "[proposed claim as sentence]" --collection {vocabulary.notes_collection} -n 5
Get timestamp
Mark extract task done (replace TASK_ID with actual task ID)What does the reduce skill do?
Extract structured knowledge from source material. Comprehensive extraction is the default — every insight that serves the domain gets extracted. For domain-relevant sources, skip rate must be below 10%. Zero extraction from a domain-relevant source is a BUG. Triggers on "/reduce", "/reduce [file]", "extract insights", "mine this", "process this".
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill reduce --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.