daily-hunt

Name: daily-hunt
Author: MagicCube

489stars57forks

How to install daily-hunt?

magiccube/agentara/daily-hunt

$npx -y skills add MagicCube/agentara --skill daily-hunt

Installs into the current project.

Summary

Daily tech product and open-source discovery digest. Fetches today's Product Hunt leaderboard and GitHub Trending top 10, then curates a personalized briefing in the user's preferred language using their memory/profile to highlight items most relevant to them. Use this skill when

Files · 1

View on GitHub

SKILL.md

1	# Daily Hunt 🏹
2
3	A personalized daily tech discovery digest combining Product Hunt launches and GitHub Trending repositories, tailored to the user's interests and expertise.
4
5	## Workflow
6
7	### Step 1: Determine Date & Data Sources
8
9	1. Note today's date.
10	2. Product Hunt launches happen Mon–Fri (Pacific Time). If today is Saturday/Sunday, use the most recent Friday's leaderboard instead and note this in the output.
11	3. Construct the PH leaderboard URL: `https://www.producthunt.com/leaderboard/daily/{YYYY}/{M}/{D}`
12
13	### Step 2: Fetch Product Hunt Data
14
15	Use `web_search` and `web_fetch` to gather PH daily leaderboard data:
16
17	1. Primary: Fetch `https://www.producthunt.com/leaderboard/daily/{YYYY}/{M}/{D}`
18	2. Backup: Search `hunted.space history {month} {year}` for structured launch data with upvotes
19	3. Backup 2: Fetch `https://hunted.space/stats` for recent daily stats
20	4. Try to get at least the top 5-10 launches with:
21	- Product name
22	- Tagline / one-line description
23	- Category tags
24	- Upvote count (if available)
25	- Comment count (if available)
26	5. For each product, construct its PH link: `https://www.producthunt.com/posts/{product-slug}`
27	- If the exact slug is unknown, link to the daily leaderboard page instead
28
29	### Step 3: Fetch GitHub Trending Data
30
31	1. Primary: Use `web_fetch` on `https://github.com/trending` to get today's trending repos
32	2. Extract top 10 repositories with:
33	- Repository full name (owner/repo)
34	- Description
35	- Primary language
36	- Star count (total)
37	- Fork count
38	- Stars gained today
39	3. Construct links: `https://github.com/{owner}/{repo}`
40
41	### Step 4: Personalize & Curate
42
43	Cross-reference the fetched data with the user's profile from memory (userMemories):
44
45	1. Identify relevance signals — Check each item against the user's:
46	- Tech stack (languages, frameworks, tools)
47	- Work domain (e.g., medical tech, AI agents, developer tools)
48	- Side projects and interests (e.g., agent architecture, AIGC, photography)
49	- Favorite topics (e.g., open source, LangChain, MCP, TypeScript)
50
51	2. Mark noteworthy items with emoji indicators:
52	- ⭐ — Directly relevant to the user's tech stack or active projects
53	- 🔥 — Exceptionally hot / high engagement / viral
54	- 👀 — Interesting for the user's broader interests or domain
55	- No emoji — Standard inclusion, still worth knowing about
56
57	3. Preserve original ranking — Keep the exact order from the source (PH leaderboard rank, GitHub Trending page order). Do NOT re-sort by engagement or relevance. Only add emoji indicators (⭐🔥👀) to mark noteworthy items in-place.
58
59	### Step 5: Format Output
60
61	Use the user's preferred language (check memory — if Chinese-speaking, use Chinese with full-width punctuation).
62
63	#### Output Structure
64
65	Use list format (not tables) for all items — this renders correctly on mobile and in IM apps.
66
67	```
68	# Daily Hunt 🏹 — {date}
69
70	> {One-sentence personalized summary of today's highlights}
71
72	## Product Hunt 热门发布
73
74	1. ⭐ [Name](link)
75	tagline
76	→ personalized reason why it matters
77
78	2. [Name](link)
79	tagline
80	→ reason
81
82	...
83
84	---
85
86	## GitHub Trending Top 10
87
88	1. 🔥 [owner/repo](link) · Python · ⭐31.4k · 🍴2.1k · +1,122 今日
89	description
90
91	2. ⭐ [owner/repo](link) · TypeScript · ⭐8.9k · 🍴430 · +300 今日
92	description
93
94	...
95
96	## 📌 昕哥精选
97
98	{2-3 sentences highlighting the most relevant items for this specific user,
99	explaining WHY these matter for their work/projects. Reference specific
100	projects or interests from memory.}
101	```
102
103	### Key Rules
104
105	- Use lists, not tables — tables break on mobile and in most IM apps
106	- Product Hunt items MUST include: name, link, tagline, personalized "why care" note (prefixed with →)
107	- GitHub items MUST include: full repo name, link, language, star count, fork count, today's stars, description
108	- "Today's stars" (stars gained today) should be shown when available from the trending page
109	- If data for a source is partially unavailable, still show what you can and note the gap
110	- The "Personal Picks" section at the end is the differentiator — make it genuinely insightful
111	- Do NOT pad with generic descriptions —

People who liked this also viewed

6 picks

Type

daily-hunt

489stars57forks

View on GitHub

How to install daily-hunt?

magiccube/agentara/daily-hunt

$npx -y skills add MagicCube/agentara --skill daily-hunt

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Daily Hunt 🏹
2
3	A personalized daily tech discovery digest combining Product Hunt launches and GitHub Trending repositories, tailored to the user's interests and expertise.
4
5	## Workflow
6
7	### Step 1: Determine Date & Data Sources
8
9	1. Note today's date.
10	2. Product Hunt launches happen Mon–Fri (Pacific Time). If today is Saturday/Sunday, use the most recent Friday's leaderboard instead and note this in the output.
11	3. Construct the PH leaderboard URL: `https://www.producthunt.com/leaderboard/daily/{YYYY}/{M}/{D}`
12
13	### Step 2: Fetch Product Hunt Data
14
15	Use `web_search` and `web_fetch` to gather PH daily leaderboard data:
16
17	1. Primary: Fetch `https://www.producthunt.com/leaderboard/daily/{YYYY}/{M}/{D}`
18	2. Backup: Search `hunted.space history {month} {year}` for structured launch data with upvotes
19	3. Backup 2: Fetch `https://hunted.space/stats` for recent daily stats
20	4. Try to get at least the top 5-10 launches with:
21	- Product name
22	- Tagline / one-line description
23	- Category tags
24	- Upvote count (if available)
25	- Comment count (if available)
26	5. For each product, construct its PH link: `https://www.producthunt.com/posts/{product-slug}`
27	- If the exact slug is unknown, link to the daily leaderboard page instead
28
29	### Step 3: Fetch GitHub Trending Data
30
31	1. Primary: Use `web_fetch` on `https://github.com/trending` to get today's trending repos
32	2. Extract top 10 repositories with:
33	- Repository full name (owner/repo)
34	- Description
35	- Primary language
36	- Star count (total)
37	- Fork count
38	- Stars gained today
39	3. Construct links: `https://github.com/{owner}/{repo}`
40
41	### Step 4: Personalize & Curate
42
43	Cross-reference the fetched data with the user's profile from memory (userMemories):
44
45	1. Identify relevance signals — Check each item against the user's:
46	- Tech stack (languages, frameworks, tools)
47	- Work domain (e.g., medical tech, AI agents, developer tools)
48	- Side projects and interests (e.g., agent architecture, AIGC, photography)
49	- Favorite topics (e.g., open source, LangChain, MCP, TypeScript)
50
51	2. Mark noteworthy items with emoji indicators:
52	- ⭐ — Directly relevant to the user's tech stack or active projects
53	- 🔥 — Exceptionally hot / high engagement / viral
54	- 👀 — Interesting for the user's broader interests or domain
55	- No emoji — Standard inclusion, still worth knowing about
56
57	3. Preserve original ranking — Keep the exact order from the source (PH leaderboard rank, GitHub Trending page order). Do NOT re-sort by engagement or relevance. Only add emoji indicators (⭐🔥👀) to mark noteworthy items in-place.
58
59	### Step 5: Format Output
60
61	Use the user's preferred language (check memory — if Chinese-speaking, use Chinese with full-width punctuation).
62
63	#### Output Structure
64
65	Use list format (not tables) for all items — this renders correctly on mobile and in IM apps.
66
67	```
68	# Daily Hunt 🏹 — {date}
69
70	> {One-sentence personalized summary of today's highlights}
71
72	## Product Hunt 热门发布
73
74	1. ⭐ [Name](link)
75	tagline
76	→ personalized reason why it matters
77
78	2. [Name](link)
79	tagline
80	→ reason
81
82	...
83
84	---
85
86	## GitHub Trending Top 10
87
88	1. 🔥 [owner/repo](link) · Python · ⭐31.4k · 🍴2.1k · +1,122 今日
89	description
90
91	2. ⭐ [owner/repo](link) · TypeScript · ⭐8.9k · 🍴430 · +300 今日
92	description
93
94	...
95
96	## 📌 昕哥精选
97
98	{2-3 sentences highlighting the most relevant items for this specific user,
99	explaining WHY these matter for their work/projects. Reference specific
100	projects or interests from memory.}
101	```
102
103	### Key Rules
104
105	- Use lists, not tables — tables break on mobile and in most IM apps
106	- Product Hunt items MUST include: name, link, tagline, personalized "why care" note (prefixed with →)
107	- GitHub items MUST include: full repo name, link, language, star count, fork count, today's stars, description
108	- "Today's stars" (stars gained today) should be shown when available from the trending page
109	- If data for a source is partially unavailable, still show what you can and note the gap
110	- The "Personal Picks" section at the end is the differentiator — make it genuinely insightful
111	- Do NOT pad with generic descriptions —

People who liked this also viewed

6 picks

Type