src/built-in/chat/skills/exhaustive-summary/SKILL.md
Summarize every file in a folder or the whole workspace exhaustively — reads each file individually, writes a per-file summary, then combines them. Use when the user asks for a comprehensive overview that must not skip files ("summarize everything in X", "what does each file in Y do"). Not for a structural-only overview (use folder-overview) or for comparing documents (use document-comparison).
npx skillsauth add mufarosean/parallx exhaustive-summaryInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
You are executing the exhaustive-summary skill. Follow these steps precisely. Do not skip any step. Do not summarize from memory — read every file.
Use list_files to enumerate every file in the target scope ($ARGUMENTS or
the entire workspace root if no scope is specified).
For every file in the coverage checklist:
read_file to read the file's full content.Do NOT skip files. Do NOT say a file is "too large to read." Every file must be read and summarized.
If a file is very short (< 3 lines), note it as a stub and describe what little content it has.
If a file contains irrelevant content (personal notes, off-topic material), still summarize it but note that it appears unrelated to the workspace's main topic.
Present the summaries in a structured format:
Compare your summary list against the coverage checklist from Step 1. If any file is missing from your summary, go back and read it now.
State: "Coverage: X/Y files summarized" where X = files you summarized and Y = total files from Step 1. X must equal Y.
testing
Extract one specific fact, field, or value from every file in a scope (folder or workspace) and aggregate the results into a structured table. Use when the user asks "what is X in each of these files", "pull all the Y values from folder Z". Not for summarizing files (use exhaustive-summary) or for comparing a small set of named documents (use document-comparison).
testing
Produce a structural overview of a folder — file count, file types, folder hierarchy, and a one-line purpose per file. Use when the user asks how a folder is organized or what's in it at a glance, without needing deep content. For per-file detailed summaries use exhaustive-summary; for extracting specific facts use scoped-extraction.
testing
Compare two or more documents side-by-side across structure, content, claims, and data, producing a structured diff of similarities, differences, and contradictions. Use when the user asks to compare, contrast, diff, or reconcile two or more named documents. Not for summarizing a single document (use exhaustive-summary) or for extracting one specific field across many files (use scoped-extraction).
documentation
Fetch GitHub issues, spawn sub-agents to implement fixes and open PRs, then monitor and address PR review comments. Usage: /gh-issues [owner/repo] [--label bug] [--limit 5] [--milestone v1.0] [--assignee @me] [--fork user/repo] [--watch] [--interval 5] [--reviews-only] [--cron] [--dry-run] [--model glm-5] [--notify-channel -1002381931352]