Thorough digest
Exhaustively review local materials with parallel sub-agents so no item is skipped. USE FOR: batch process local files, digest a folder, synthesize narrative from existing materials. DO NOT USE FOR: web research, single-document summaries, slide or report writing.From its SKILL.md
npx -y skills add specter119/skills --skill thorough-digestAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
2.3 KB, as published. Nobody here has run it
Thorough Digest
Responsible for exhaustive processing and parallel synthesis of local materials; does not handle external web research.
USE FOR
- Batch processing of local files, directories, sections, or material collections
- Item-by-item extraction of key points with no items allowed to be skipped
- Building a unified narrative or comprehensive report from existing materials
DO NOT USE FOR
- Requires fetching new information from the web
- Only processing one or two documents
- The primary goal is writing slides or a formal report rather than first digesting materials
Execution Skeleton
- First inventory all input items to ensure the scope is clear.
- Follow workflow to determine grouping and parallelization strategy.
- Give each sub-agent an explicit item list, output path, and the constraint that no items may be skipped.
- Aggregate findings from each group; if needed, chain an external research skill to fill gaps.
Reference Map
- workflow: phased workflow, grouping strategy, output structure
- guidelines: parallelization granularity, error handling, best practices
- templates: inventory / classify / sub-agent / synthesis templates
- trigger-cases: minimal trigger examples
- execution-cases: key execution scenarios
- optimization-notes: current refactor rationale
Output Contract
- Default deliverables: inventory, grouping results, per-group findings, and final synthesis
- Must be able to report total input count, processed count, and coverage status
Collaboration and Handoff
- If the synthesis phase reveals critical fact gaps, hand off to an external research skill
- If materials are fully digested and the next step is writing a formal deliverable, hand off to a writing or presentation-structure skill