From cascade-mcp
Summarizes and categorizes content from Jira issues, Confluence pages, and Google Docs in .temp/cascade/context/. Extracts key details like requirements, links, criteria; discovers new URLs and writes summary files.
npx claudepluginhub bitovi/cascade-mcp --plugin cascade-mcpThis skill uses the workspace's default tool permissions.
Summarize and categorize raw content previously fetched by the `load-linked-resource-content` sub-skill. Extract key information and discover any additional links embedded in the content.
Fetches raw content from Jira issues, Confluence pages, Google Docs, and Sheets via MCP tool. Saves markdown files locally and appends discovered links to to-load.md for iterative loading.
Generates executive summaries from date-folder communication files (MD/JSON/TXT/CSV) using AI with automatic workspace context from static/ PROJECT/TEAM files.
Ingests raw source documents into structured wiki pages for sources, entities, and concepts with summaries, key claims, and bidirectional wikilinks. Use for building knowledge bases from articles or notes.
Share bugs, ideas, or general feedback.
Summarize and categorize raw content previously fetched by the load-linked-resource-content sub-skill. Extract key information and discover any additional links embedded in the content.
This is a sub-skill β called by parent skills after load-linked-resource-content has fetched raw content. Use when the parent skill needs summarized, structured content before synthesis or generation steps.
.temp/cascade/context/ (written by load-linked-resource-content).temp/cascade/context/to-load.md exists with loading manifestRead all .md files in .temp/cascade/context/ (excluding to-load.md and any files ending in -summary.md).
For each raw content file, produce a summary that extracts:
For Jira issues (jira-*.md):
For Confluence pages (confluence-*.md):
For Google Docs (gdoc-*.md):
Save each summary alongside the raw content file with a -summary.md suffix:
.temp/cascade/context/
βββ jira-PROJ-123.md β raw content (from load-linked-resource-content)
βββ jira-PROJ-123-summary.md β summary (from summarize-document-content)
βββ confluence-design-spec.md
βββ confluence-design-spec-summary.md
βββ gdoc-requirements.md
βββ gdoc-requirements-summary.md
βββ to-load.md
While analyzing, collect any URLs found in the content that are NOT already in to-load.md. Append them as [ ] entries in the ## Unloaded section.
Report back:
to-load.md)Keep summaries concise but complete. Target 200-500 words per file. Use this structure:
# Summary: {source title}
**Source**: {URL}
**Type**: {Jira Issue | Confluence Page | Google Doc}
## Key Information
- {bullet points of the most important facts}
## Requirements / Specifications
- {requirements extracted from the content}
## Open Questions
- β {any unanswered questions found}
- π¬ {any answered questions found with their answers}
## Discovered Links
- {new URLs found in this content}
-summary.md β skip them unless the parent skill explicitly requests re-analysis