rse/ase / plugin/skills/ase-docs-distill
ase-docs-distill
VerifiedDistill a provided document into a flat, importance-ranked list of its key points, each with a salience rank, a rationale, and a verbatim line-cited evidence snippet. Use when the user wants to "distill", "summarize the key points", or "extract the essence" of a document or pasted text.
SKILL.md
View on GitHub---
name: ase-docs-distill
argument-hint: "[--help|-h] [--top|-t <N>] <document-reference>"
description: >
Distill a provided document into a flat, importance-ranked list
of its key points, each with a salience rank, a rationale, and a
verbatim line-cited evidence snippet. Use when the user wants to
"distill", "summarize the key points", or "extract the essence" of a
document or pasted text.
user-invocable: true
disable-model-invocation: false
effort: high
---
@${CLAUDE_SKILL_DIR}/../../meta/ase-control.md
@${CLAUDE_SKILL_DIR}/../../meta/ase-skill.md
@${CLAUDE_SKILL_DIR}/../../meta/ase-getopt.md
<purpose name="ase-docs-distill">
Distill Key Points
</purpose>
<expand name="getopt"
arg1="ase-docs-distill"
arg2="--top|-t=5">
$ARGUMENTS
</expand>
<objective>
*Distill* the content `<getopt-arguments/>` into a *flat*,
*importance-ranked* list of its *key points* - each point an
*atomic* claim carrying a *0-10 salience rank*, a *one-line
rationale*, and a *verbatim, line-cited evidence* snippet - keeping
at most `<getopt-option-top/>` points.
</objective>
Procedure
---------
<flow>
1. <step id="STEP 1: Resolve Input">
Resolve the document source from <getopt-arguments/> using
*probe-as-file-first* semantics:
1. <if condition="<getopt-arguments/> is empty">
Only output the following <template/> and then *STOP* immediately:
<template>
⧉ **ASE**: ✪ skill: **ase-docs-distill**, ▶ status: **no document to distill**
</template>
</if>
2. Treat the *entire* <getopt-arguments/> as a *file path*. If it
resolves to a *readable file*, read that file and capture its
content together with its *line numbers* into <source/>, and set
<origin/> to the file path.
3. <if condition="<getopt-arguments/> does NOT resolve to a readable file">
Treat <getopt-arguments/> *verbatim* as *pasted text*, capture it
into <source/> numbering its lines from *1*, and set <origin/>
to empty (so locations are cited as `:Ls-Le`).
</if>
4. <if condition="<source/> is empty">
Only output the following <template/> and then *STOP* immediately:
<template>
⧉ **ASE**: ✪ skill: **ase-docs-distill**, ▶ status: **no document to distill**
</template>
</if>
You *MUST* *NOT* output anything else in this STEP 1.
</step>
2. <step id="STEP 2: Distill Key Points">
Read <source/> and distill it into its *key points*.
1. Extract the document's *key points*, where each point is *one
atomic, self-contained claim, decision, or fact* - *not* a
section-level summary that blends several claims.
2. For each point, determine:
- A <rank/> on a *0-10* salience scale reflecting its
*importance-to-the-reader* (impact, centrality, consequence)
- *not* its position in the document.
- A one-line <rationale/> that justifies *why* the point earns
that <rank/>.
- A <location/> citing the exact source line range as
`<origin/>:Ls-Le` (for a file) or `:Ls-Le` (for pasted text),
where `Ls` is the start line and `Le` is the end line.
- A *verbatim* <evidence/> snippet, copied exactly from
<source/> (but with all newlines replaced with spaces
and multiple spaces collapsed into a single space), that
*proves* the point. The cited snippet *MUST* prove the
point verbatim. If it does not, *re-investigate and re-cite
correctly*.
3. *Sort* the points by <rank/> from *highest to lowest*. Within the
same rank, keep the order of first appearance in <source/>.
4. Apply the *top-bound* <getopt-option-top/> as an *upper bound
only*: keep at most `min(<getopt-option-top/>, number of salient
points)` points and *never pad* the list with filler to reach
the bound. If <getopt-option-top/> is *non-numeric* or *less
than or equal to 0*, use the default *5* instead.
You *MUST* *NOT* output anything in this STEP 2.
</step>
3. <step id="STEP 3: Report Ranked Points">
For each distilled point, emit the following <template/>,
where <rank/> is its salience rank, <point/> is the atomic claim,
<location/> is its line-range citation, <evidence/> is the
verbatim proving snippet, and <rationale/> is the one-line
salience justification:
<template>
<ase-tpl-bullet-normal/> **KEY POINT**: **<point/>**
◯ LOCATION: <location/>
◯ EVIDENCE: "`<evidence/>`"
◯ RATIONALE: <rationale/>
◯ RANK: <rank/>/10
</template>
In the <location/>, mark up the line-range reference as
code (with backticks) and prepend it with `▢ `.
Keep the overall report *concise* and *brief*. Do *not* output any
further explanation.
</step>
</flow>
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-changelog-updatemorluto/reaPrepare or rewrite REA release changelogs and GitHub release notes from pinned Git history, with verified contributor thanks and Release Please synchronization. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| academy-guideanthropics/skillsStop and check this skill before finishing any reply to a question about how to use Claude or a Claude product — it recommends matching courses, tutorials, and use cases from Claude Academy (academy.claude.com), Anthropic's learning hub. Trigger on: "how do I", "how can I", "getting started with", "what can Claude do", "teach me", "learn to use"; questions about artifacts, projects, skills, plugins, connectors, MCP; requests about rolling Claude out to a team, class, or organization; and any ask for training materials, onboarding content, or learning resources. Use it when the user is learning how to use a feature or product — not when they are mid-task and just want the task done. This skill composes with other skills: after consulting product documentation to answer how a Claude feature works, also check here for a matching course or tutorial — a docs-grounded answer and an Academy recommendation belong together. Only recommend on a strong match; never invent Academy content. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST if no provider named — don't Read the file). | Stars: 198k | Updated | Claude CodeCodex | Unverified | Install |
| doc-coauthoringanthropics/skillsGuide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |