Extract atomic claims with provenance, confidence, and topic tags from unstructured documents.
npx skills add https://github.com/kinnerkarmanish/mak --skill reduce-kinnerkarmanish
Please help me install this Agent Skill. Skill: reduce Source: https://github.com/kinnerkarmanish/mak/tree/main/library/skills/processing/reduce Command: npx skills add https://github.com/kinnerkarmanish/mak --skill reduce-kinnerkarmanish
Transform raw input (articles, transcripts, notes, documents) into structured atomic claims — each with provenance, confidence, and topic tags. This is the core extraction step: it turns noise into signal and enables recombination of individual claims.
Instruct the system to reduce a source document to a set of atomic, provenance-rich claims.
High-intent search queries and answers about installing and using this skill.
Extracting atomic claims from unstructured articles and transcripts involves processing raw text into structured statements. This Skill automates that extraction, attaching provenance, confidence, and topic tags to each claim for reuse.
The best way to add provenance metadata to extracted text claims is by generating YAML frontmatter. This format structures source metadata, claim details, and confidence levels directly with the extracted data.
Yes, you can process interview transcripts into structured data with confidence scores. The system reduces long transcripts into compact sets of actionable claims, each tagged with provenance and confidence metrics.
Yes, the claim extraction process supports depth and format options for documents. It satisfies deterministic input processing, allowing you to specify the extraction depth when reducing source material.
Turning raw notes into structured, reusable statements for data analytics requires reducing the text into atomic claims. This transforms unstructured noise into signal by outputting provenance-rich statements with topic tags.
Using atomic claims instead of summarizing the entire source document enables traceability and recombination. Each individual claim retains its specific provenance and confidence level, whereas summaries lose track of distinct facts.
dkyazzentwatwaFormat raw meeting notes into structured Markdown or PDF documents.
MantisWareExtract atomic claims with provenance from unstructured text sources.
hjemmesidekongenExtract atomic instincts from JSONL observations and update YAML definitions.
luispabonCondense long-form text into structured summaries and actionable takeaways.
