read-content-digest

Name: read-content-digest
Author: SpaceZephyr

56stars18forks

How to install read-content-digest?

spacezephyr/read-buddy/read-content-digest

$npx -y skills add SpaceZephyr/read-buddy --skill read-content-digest

Installs into the current project.

Summary

Transform long-form content (YouTube videos, podcasts, interviews, articles) into engaging short-form and long-form narratives. Extracts core insights and presents them in two styles: concise social media posts (300-800 characters with numbered emoji lists) and detailed narrative

Files · 1

View on GitHub

SKILL.md

1	# Content Digest
2
3	Transform long-form content into compelling short-form and long-form narratives.
4
5	## Overview
6
7	This skill converts lengthy content (YouTube videos, podcasts, interviews, articles) into two distinct formats:
8
9	1. Short-Form (短文): Social media-friendly summaries (300-800 characters) with numbered emoji lists (1️⃣2️⃣3️⃣)
10	2. Long-Form (长文): Narrative articles (1500-3000+ characters) with story arcs, section headers, and integrated quotes
11
12	## Workflow
13
14	### 1. Obtain the Content
15
16	If user provides a URL:
17	- YouTube links: Use WebFetch or attempt to extract transcript
18	- Article URLs: Use WebFetch to retrieve content
19	- Podcast links: Fetch transcript if available
20
21	If user provides text:
22	- Read the full transcript or article text directly
23
24	If content is unclear:
25	- Ask: "Please provide the YouTube link, podcast transcript, or article you'd like me to transform."
26
27	### 2. Determine Output Format
28
29	If user specifies format:
30	- Proceed with their choice (short-form only, long-form only, or both)
31
32	If user does not specify:
33	- Ask: "Would you like: (1) Short-form only, (2) Long-form only, or (3) Both versions?"
34
35	Default behavior:
36	- Generate both versions to maximize value
37
38	### 3. Deep Analysis - Four-Stage Process
39
40	CRITICAL: Follow this systematic process to ensure depth
41
42	#### Stage 1: Extract All Viewpoints (50+ minimum)
43
44	Read the entire content thoroughly and extract ALL viewpoints, including:
45	- Explicit statements and opinions
46	- Implicit beliefs revealed through stories
47	- Decision-making rationales
48	- Observations about the industry/domain
49	- Personal experiences and lessons
50	- Counterexamples and contrasts
51	- Numbers, data points, specific examples
52
53	Goal: Create a comprehensive list of 50+ viewpoints before filtering. Don't judge quality yet - just extract everything.
54
55	#### Stage 2: Filter for Non-Consensus & Depth
56
57	From the 50+ viewpoints, identify and mark those that are:
58	- Non-consensus (非共识): Challenges industry conventional wisdom
59	- Personal/private insights (个人私下表达): Things people think but rarely say publicly
60	- Counterintuitive (反直觉): Surprises even informed readers
61	- Interesting trivia (有意思的冷知识): Specific details that reveal deeper patterns
62	- Mental models: Frameworks that explain decision-making
63	- Second-order insights: Not just "what" but "why this matters philosophically"
64	- Paradoxes and tensions: Contradictions that expose underlying principles
65
66	Goal: Flag the 20-30 viewpoints that pass the "non-obvious test" - would a smart, informed reader already know this?
67
68	#### Stage 3: Select Core Narrative Elements
69
70	Identify:
71	- Core narrative: What's the main story or theme?
72	- Memorable quotes: Direct quotes that capture big ideas or reveal character
73	- Turning points: Moments of realization or paradigm shifts
74	- Dramatic elements: Irony, contrast, or unexpected outcomes
75	- Specific details: Names, numbers, dates that prove the deeper point
76
77	#### Stage 4: Curate Final Insights
78
79	From the filtered viewpoints (Stage 2) and narrative elements (Stage 3):
80	- For short-form: Select 10-15 most profound, actionable insights
81	- For long-form: Use the same 10-15 insights as the foundation, then weave in narrative arc
82
83	### 4. Generate Short-Form Version
84
85	CRITICAL: Use ONLY the 10-15 curated insights from Stage 4
86
87	Consult [style-guide.md](references/style-guide.md) for detailed guidelines. See [examples.md](references/examples.md) for reference.
88
89	Structure:
90	```
91	# MMDD：[嘉宾名] X [栏目名]：[一句话核心观点]
92
93	今天看到 [嘉宾名] 去了 [栏目名] 的播客。
94
95	[嘉宾名] [2-3句话介绍嘉宾背景和核心成就，用具体数据]。
96
97	这期播客总共录了 [时长]，[嘉宾名] 谈到了 [N] 个有趣的观点：
98
99	1、[观点标题/关键词]。[完整的逻辑阐述，包含推理过程，2-4句话]
100
101	2、[观点标题/关键词]。[完整的逻辑阐述，包含推理过程，2-4句话]
102
103	3、[观点标题/关键词]。[完整的逻辑阐述，包含推理过程，2-4句话]
104
105	...
106	[Continue with 10-15 total points]
107
108	---
109	```
110
111	标题格式 (重要):
112	- 格式：`# MMDD：嘉宾名 X 栏目名：一句话观点`
113	- 示例：`# 0130：Peter Steinberger X The Pragmatic Engineer：一天600次提交，代码比以前更好`
114	- 一句话观点要抓住最反直觉或最有冲击力的点
115
116	开头格式 (重要):
117	- 第一句：简单说看了什么 `今天看到 [嘉宾] 去了 [栏目] 的播客。`
118	- 第二段：2-3句介绍嘉宾背景，必须有具体数据（数字、公司名、产品名）
119	- 第三句：过渡句 `这期播客总共录了 [时长]，[嘉宾] 谈到了 [N] 个有趣的观点：`
120	- 如果不知道时长，可以写"将近两小时"或省略时长
121
122	List Format Rules (重要):
123	- 使用数字+顿号格式 (`1、` `2、` `3、`...)
124	- 注意是顿号「、」不是点号「.」
125	- 每条是完整段落，包含：
126	- 观点/结论（第一句）
127	- 逻辑推理/因果解释（后续句子）
128	- 具体例子或数据（如有）
129	- 每条 2-4 句话，50-150 字
130	- 观点要有深度：
131	- ✅ "AI 应用创业者不会相信 AGI。逻辑很简单，如果真信，就不应该做 AI 应用创业。AGI 如果存在，创业就只剩一件事：去做有机会达成 AGI 的模型。"
132	- ❌ "AI 应用创业者不相信 AGI"（太简单，没有逻辑推理）
133	- **保持 10-15

People who liked this also viewed

6 picks

Type

read-content-digest

56stars18forks

View on GitHub

How to install read-content-digest?

spacezephyr/read-buddy/read-content-digest

$npx -y skills add SpaceZephyr/read-buddy --skill read-content-digest

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Content Digest
2
3	Transform long-form content into compelling short-form and long-form narratives.
4
5	## Overview
6
7	This skill converts lengthy content (YouTube videos, podcasts, interviews, articles) into two distinct formats:
8
9	1. Short-Form (短文): Social media-friendly summaries (300-800 characters) with numbered emoji lists (1️⃣2️⃣3️⃣)
10	2. Long-Form (长文): Narrative articles (1500-3000+ characters) with story arcs, section headers, and integrated quotes
11
12	## Workflow
13
14	### 1. Obtain the Content
15
16	If user provides a URL:
17	- YouTube links: Use WebFetch or attempt to extract transcript
18	- Article URLs: Use WebFetch to retrieve content
19	- Podcast links: Fetch transcript if available
20
21	If user provides text:
22	- Read the full transcript or article text directly
23
24	If content is unclear:
25	- Ask: "Please provide the YouTube link, podcast transcript, or article you'd like me to transform."
26
27	### 2. Determine Output Format
28
29	If user specifies format:
30	- Proceed with their choice (short-form only, long-form only, or both)
31
32	If user does not specify:
33	- Ask: "Would you like: (1) Short-form only, (2) Long-form only, or (3) Both versions?"
34
35	Default behavior:
36	- Generate both versions to maximize value
37
38	### 3. Deep Analysis - Four-Stage Process
39
40	CRITICAL: Follow this systematic process to ensure depth
41
42	#### Stage 1: Extract All Viewpoints (50+ minimum)
43
44	Read the entire content thoroughly and extract ALL viewpoints, including:
45	- Explicit statements and opinions
46	- Implicit beliefs revealed through stories
47	- Decision-making rationales
48	- Observations about the industry/domain
49	- Personal experiences and lessons
50	- Counterexamples and contrasts
51	- Numbers, data points, specific examples
52
53	Goal: Create a comprehensive list of 50+ viewpoints before filtering. Don't judge quality yet - just extract everything.
54
55	#### Stage 2: Filter for Non-Consensus & Depth
56
57	From the 50+ viewpoints, identify and mark those that are:
58	- Non-consensus (非共识): Challenges industry conventional wisdom
59	- Personal/private insights (个人私下表达): Things people think but rarely say publicly
60	- Counterintuitive (反直觉): Surprises even informed readers
61	- Interesting trivia (有意思的冷知识): Specific details that reveal deeper patterns
62	- Mental models: Frameworks that explain decision-making
63	- Second-order insights: Not just "what" but "why this matters philosophically"
64	- Paradoxes and tensions: Contradictions that expose underlying principles
65
66	Goal: Flag the 20-30 viewpoints that pass the "non-obvious test" - would a smart, informed reader already know this?
67
68	#### Stage 3: Select Core Narrative Elements
69
70	Identify:
71	- Core narrative: What's the main story or theme?
72	- Memorable quotes: Direct quotes that capture big ideas or reveal character
73	- Turning points: Moments of realization or paradigm shifts
74	- Dramatic elements: Irony, contrast, or unexpected outcomes
75	- Specific details: Names, numbers, dates that prove the deeper point
76
77	#### Stage 4: Curate Final Insights
78
79	From the filtered viewpoints (Stage 2) and narrative elements (Stage 3):
80	- For short-form: Select 10-15 most profound, actionable insights
81	- For long-form: Use the same 10-15 insights as the foundation, then weave in narrative arc
82
83	### 4. Generate Short-Form Version
84
85	CRITICAL: Use ONLY the 10-15 curated insights from Stage 4
86
87	Consult [style-guide.md](references/style-guide.md) for detailed guidelines. See [examples.md](references/examples.md) for reference.
88
89	Structure:
90	```
91	# MMDD：[嘉宾名] X [栏目名]：[一句话核心观点]
92
93	今天看到 [嘉宾名] 去了 [栏目名] 的播客。
94
95	[嘉宾名] [2-3句话介绍嘉宾背景和核心成就，用具体数据]。
96
97	这期播客总共录了 [时长]，[嘉宾名] 谈到了 [N] 个有趣的观点：
98
99	1、[观点标题/关键词]。[完整的逻辑阐述，包含推理过程，2-4句话]
100
101	2、[观点标题/关键词]。[完整的逻辑阐述，包含推理过程，2-4句话]
102
103	3、[观点标题/关键词]。[完整的逻辑阐述，包含推理过程，2-4句话]
104
105	...
106	[Continue with 10-15 total points]
107
108	---
109	```
110
111	标题格式 (重要):
112	- 格式：`# MMDD：嘉宾名 X 栏目名：一句话观点`
113	- 示例：`# 0130：Peter Steinberger X The Pragmatic Engineer：一天600次提交，代码比以前更好`
114	- 一句话观点要抓住最反直觉或最有冲击力的点
115
116	开头格式 (重要):
117	- 第一句：简单说看了什么 `今天看到 [嘉宾] 去了 [栏目] 的播客。`
118	- 第二段：2-3句介绍嘉宾背景，必须有具体数据（数字、公司名、产品名）
119	- 第三句：过渡句 `这期播客总共录了 [时长]，[嘉宾] 谈到了 [N] 个有趣的观点：`
120	- 如果不知道时长，可以写"将近两小时"或省略时长
121
122	List Format Rules (重要):
123	- 使用数字+顿号格式 (`1、` `2、` `3、`...)
124	- 注意是顿号「、」不是点号「.」
125	- 每条是完整段落，包含：
126	- 观点/结论（第一句）
127	- 逻辑推理/因果解释（后续句子）
128	- 具体例子或数据（如有）
129	- 每条 2-4 句话，50-150 字
130	- 观点要有深度：
131	- ✅ "AI 应用创业者不会相信 AGI。逻辑很简单，如果真信，就不应该做 AI 应用创业。AGI 如果存在，创业就只剩一件事：去做有机会达成 AGI 的模型。"
132	- ❌ "AI 应用创业者不相信 AGI"（太简单，没有逻辑推理）
133	- **保持 10-15

People who liked this also viewed

6 picks

Type