Stats
Actions
Tags
From slack-forge
Guidance for extracting durable organizational knowledge from local Slack transcript files.
How this skill is triggered — by the user, by Claude, or both
Slash command
/slack-forge:knowledge-harvesterThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill when analyzing local transcript snapshots to identify preservable knowledge.
Use this skill when analyzing local transcript snapshots to identify preservable knowledge.
slack-forge/transcripts/*.md.Keep items likely useful in 2+ weeks.
high: explicit decision/announcementmedium: strong contextual insightlow: tentative/speculative statementsFor each extracted knowledge item, preserve:
Content must include both the knowledge fact AND its strategic significance.
Minimum output: 1 summary paragraph + 1 significance paragraph (prefixed with **Significance:** ). A single sentence is never sufficient.
Include a direct quote from the transcript where one exists — quotes make knowledge items credible and traceable during review.
Tags must start with a memory-hint destination tag as the first element: person, project, glossary, or general. Additional descriptive tags follow.
Use tags to signal likely destination type at promote time:
npx claudepluginhub jeremybrice/the-forge --plugin slack-forgeExtracts uncaptured knowledge (decisions, feedback, context) from the current conversation before it's lost to compaction. Use after completing a task, before switching context, or before large exploratory work.
Extracts requirements, decisions, action items, and concerns from meeting transcripts with speaker role weighting and domain organization. Processes globs and conversational formats.
Extracts key decisions, takeaways, action items with DRIs, and open questions from meeting transcripts, notes, or Slack threads. Handles single meetings or weekly batches.