Compress large Markdown files into verifiable inline digests preserving facts and IDs.
npx skills add https://github.com/pdlc-os/pdlc --skill distill-pdlc-os
Please help me install this Agent Skill. Skill: distill Source: https://github.com/pdlc-os/pdlc/tree/main/skills/distill Command: npx skills add https://github.com/pdlc-os/pdlc --skill distill-pdlc-os
Distill compresses large markdown artifacts into a dense, inline digest appended to the source file, enabling sub-agents to read a compact, fact-preserving summary without scanning the full prose.
Run the distill process on a Markdown file to generate and attach a verifiable inline digest.
High-intent search queries and answers about installing and using this skill.
To compress large Markdown documentation, you run a distillation process that appends a dense, inline digest to the source file. This digest preserves every fact, ID, path, and table, enabling sub-agents to read a compact summary without scanning full prose.
Round-trip verification is a mechanism proving that a compressed Markdown digest can reconstruct the original facts. It ensures no information is lost during compression, allowing sub-agents to safely read digests instead of full prose.
You configure distillation thresholds by setting project-specific parameters and whitelists. These settings determine when the distillation process applies the inline digest, balancing context preservation and performance for your documentation automation.
Yes, the Markdown digest tool explicitly preserves every fact, ID, file path, and table from the source file. This ensures that the generated inline digest maintains all critical metadata for round-trip reconstruction.
You should use an inline digest for Markdown files when sub-agents need to process large artifacts without scanning full prose. It provides a compact, fact-preserving summary that respects project-configured thresholds to optimize context and performance.
Yes, you can reconstruct original facts from an appended digest. The distillation process creates a verifiable inline digest designed for round-trip reconstruction, ensuring every fact, ID, and path is preserved and recoverable.
ldk-hubSplit large markdown documents into shard files by level-2 sections.
SkogAIDispatch independent tasks to parallel agents and integrate their outputs.
mwridgwayAnalyzes CS2 demo files to generate tactical markdown reports with player statistics and round breakdowns.
oimiragieoCondense long documents, code, and conversations into concise summaries.
