技能详情（站内镜像，无评论）
作者：Superlore @AdamJurgens
许可证：MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本：v1.2.0
统计：⭐ 0 · 421 · 0 current installs · 0 all-time installs
⭐ 0
安装量（当前） 0
🛡 VirusTotal ：良性 · OpenClaw ：可疑
Package：adamjurgens/openclaw-podcast
安全扫描（ClawHub）
- VirusTotal ：良性
- OpenClaw ：可疑
OpenClaw 评估
The skill largely does what it says (turn workspace content into podcast episodes via Superlore.ai), but the included scripts read/walk global home paths, can persist your API key to shell profiles, and include a configurable API URL / command-execution capability that broadens its reach beyond what's strictly described — review before installing or saving credentials.
目的
The skill's name/description match its main requirements: it needs a SUPERLORE_API_KEY and network access to the Superlore API and reads workspace files (memory/*.md, JOBS.md, etc.) to build episode prompts. However, the scripts also read USER.md, workspace fallbacks under ~/.openclaw, and global podcast-styles under ~/.openclaw — behavior not explicitly enumerated in the high-level description. The setup wizard also offers to write the API ke…
说明范围
SKILL.md and references explicitly describe sending workspace content to Superlore and scheduling via cron. The shipped scripts, however, let you override the API URL (--api-url), load global and workspace-level custom styles (~/.openclaw/podcast-styles and workspace/podcast-styles), and read USER.md — i.e., they can read files beyond the four listed memory files. The setup wizard prompts to append the API key to shell profiles and may execute…
安装机制
No install spec (instruction-only plus scripts). That minimizes risk from downloading arbitrary archives. The scripts use only Node built-ins (https, fs, path, child_process), and there is no remote code fetch in the install step.
证书
Only SUPERLORE_API_KEY is declared required, which is appropriate for a third‑party TTS/API integration. However, the setup wizard can write that key into shell RC files (persisting it on disk), and the code inspects HOME and SHELL environment variables and global config paths (~/.openclaw). Those behaviors increase the impact of granting the single API key (and persisting it broadly) and should be considered when deciding where/how to store c…
持久
The skill does not force persistence via always:true, but the setup wizard offers to persist the API key to ~/.zshrc or ~/.bashrc and emits cron commands to schedule recurring episode generation. The code also looks for global custom styles in ~/.openclaw and may execute system commands (execFileSync imported). These are legitimate setup conveniences but are higher‑privilege actions (file writes and possible command execution) that require exp…
安装（复制给龙虾 AI）
将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「Openclaw Podcast」。简介：Transform your OpenClaw workspace into personalized AI-powered podcast briefing…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/adamjurgens/openclaw-podcast/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
---
name: openclaw-podcast
description: Transform your OpenClaw workspace into personalized AI-powered podcast briefings. Get daily audio updates on your work, priorities, and strategy in 8 compelling styles—from documentary narrator to strategic advisor. Connects directly to your agent's memory and files. Includes 3 free hours of podcast generation with your Superlore.ai API key. Schedule morning, midday, or evening briefings that keep you informed without reading screens.
metadata:
openclaw:
requires:
env:
- name: SUPERLORE_API_KEY
description: "API key from superlore.ai — get one free at https://superlore.ai or use the setup wizard's email OTP flow"
required: true
permissions:
- network: "HTTPS requests to superlore-api.onrender.com (official Superlore API hosted on Render)"
- filesystem: "Reads workspace files (memory/*.md, JOBS.md, HEARTBEAT.md, MEMORY.md). Setup wizard optionally appends env var to ~/.zshrc or ~/.bashrc (user-confirmed)."
- cron: "Setup wizard outputs openclaw cron commands for scheduling. Runs them only with explicit user confirmation."
---
# openclaw-podcast
**Your AI agent already knows everything about your work. Now it can brief you on it.**
Every day, your OpenClaw agent accumulates context — your accomplishments, your metrics, your blockers, your priorities. This skill turns all of that into a professional podcast briefing via [Superlore.ai](https://superlore.ai). No forms to fill out. No manual summaries. Just press play.
**3 free hours included with your API key. Setup takes 2 minutes.**
---
## 🎧 Hear It For Yourself — 8 Sample Episodes
These are real episodes generated by this skill for fictional personas. Each one demonstrates a different briefing style. Listen and imagine: *what would this sound like if the AI knew YOUR workspace?*
---
### 🎙️ The Briefing
**Sarah Chen** — VP Engineering, TaskFlow (B2B SaaS, Series B)
> *"The latency drop from 340 milliseconds to 47 milliseconds isn't just a performance win — it's a signal that the architecture is finally doing what it was designed to do..."*
Sarah's agent reads her sprint board, PR merge history, incident logs, and engineering metrics every evening and delivers a polished documentary-style briefing — like having a personal analyst who's been embedded with her team for months.
🎧 **[Listen: The Briefing — Sarah Chen](https://superlore.ai/episode/cmlyu95qm0006mvy4gue30cj6)**
---
### 📈 Opportunities & Tactics
**Blake Rivera** — Founder, Lumière Skin (DTC Skincare, $312K/month)
> *"The subscription conversion gap is hiding $100,000 in monthly recurring revenue inside your existing customer base. And there's a spa chain inquiry sitting unanswered for 8 days..."*
Blake's agent spotted a 12% vs 18-25% industry benchmark gap in subscription conversions, calculated the revenue sitting on the table, and flagged a missed wholesale opportunity — all from reading his metrics files.
🎧 **[Listen: Opportunities & Tactics — Blake Rivera](https://superlore.ai/episode/cmlyu9al90008mvy46oxs4fvm)**
---
### 🚀 10X Thinking
**Marcus Webb** — Solo Developer, Nexus AI Framework (2,847 GitHub stars)
> *"LangChain has 80,000 stars. You have 2,847. But that might be the wrong comparison entirely. What if Nexus isn't a framework — what if it's the standard?"*
Marcus's agent reads his GitHub activity, Discord metrics, and technical notes — then challenges his assumptions about what he's actually building and who he should be competing with.
🎧 **[Listen: 10X Thinking — Marcus Webb](https://superlore.ai/episode/cmlyu9f78000amvy4i27vek90)**
---
### 🧭 The Advisor
**Priya Nair** — First-Time Founder, SkillBridge (EdTech, Pre-Seed)
> *"I notice you spent 40 hours on product this month and 8 hours on fundraising. With 7 months of runway and a Precursor pitch in 8 days, that's a pattern worth naming..."*
Priya's agent reviews her time allocation, pipeline status, and discovery call notes — then delivers the honest mentor feedback she might not be getting anywhere else.
🎧 **[Listen: The Advisor — Priya Nair](https://superlore.ai/episode/cmlyu9jic000cmvy4dzqi4fmc)**
---
### 🎯 Focus & Priorities
**Alex Kim** — Engineering Manager, FAANG-Scale ($2.4B ad revenue org)
> *"4.5 hours in meetings, 2.5 hours in Slack, 1.5 hours of actual deep work. That's not engineering leadership — that's reactive management. And two engineers are quietly making noise about leaving..."*
Alex's agent analyzes his calendar log, Slack volume, and org context — then cuts straight to what's actually important versus what just feels urgent.
🎧 **[Listen: Focus & Priorities — Alex Kim](https://superlore.ai/episode/cmlyu9o3v000emvy4qi9rgzli)**
---
### 📊 Growth & Scale
**Jordan Park** — Growth Lead, Meridian Finance (YC W23, 84K users)
> *"The cohort data is unambiguous: users who connect a second bank account within 48 hours have 3.2x better 90-day retention. That single insight is worth rebuilding your onboarding around..."*
Jordan's agent processed 30-day cohort analysis results, calculated the AARRR funnel bottlenecks, and identified the one activation metric that changes everything.
🎧 **[Listen: Growth & Scale — Jordan Park](https://superlore.ai/episode/cmlyu9scx000gmvy45wozwrq7)**
---
### 📅 Week in Review
**Elena Vasquez** — CTO, ClearPath Health (Series A, 31 hospital clients)
> *"Monday was the compliance unlock. Tuesday was the technical launch. Wednesday opened a $180K door. Thursday fixed what broke. Friday built the map. The story of this week is a team that ships fast and fixes faster..."*
Elena's agent synthesizes the week's wins, deployments, customer conversations, and metrics into a narrative arc — giving her the zoom-out perspective that's hard to find when you're in the middle of it.
🎧 **[Listen: Week in Review — Elena Vasquez](https://superlore.ai/episode/cmlyu9wx7000imvy4unlzu3gn)**
---
### 🔮 The Futurist
**David Okonkwo** — Founder, Terravolt Energy (Climate-Tech, $4.2M Seed)
> *"The 43% curtailment reduction isn't just a pilot result — it's proof that AI can unlock 85 gigawatts of stranded renewable capacity sitting idle on the US grid. That's the size of the prize you're building toward..."*
David's agent connects today's Duke Energy pilot results, provisional patent filing, and DOE grant submission to the 3/6/12-month trajectory — keeping the mission visible when the details get overwhelming.
🎧 **[Listen: The Futurist — David Okonkwo](https://superlore.ai/episode/cmlyua1c8000kmvy47l1f9b5a)**
---
## ✨ What Makes This Different
Most AI summaries are generic. They know what you typed in a chat box. This skill knows what you actually *did* — because it reads the same workspace files your agent uses every day.
**The difference in practice:**
| Generic AI Summary | OpenClaw Podcast |
|---|---|
| "You worked on some engineering tasks today" | "The latency drop from 340ms to 47ms unblocks the enterprise tier rollout you've been waiting on for two months" |
| "You should focus on your priorities" | "You spent 40 hours on product this month and 8 on fundraising — with 7 months of runway, that ratio needs to flip" |
| "Consider growth opportunities" | "Your subscription conversion at 12% vs. the 18-25% industry benchmark is $100K/month hiding in your existing customer base" |
The AI knows your project names. Your metrics. Your blockers. Your job board. The context that takes 10 minutes to explain to a consultant — your agent already has it.
**What Superlore adds on top:**
- 📝 **Notes** — Save insights from episodes to your library
- 🎵 **Episode library** — Listen back to any briefing, any time
- 🗣️ **Premium TTS voices** — Kokoro-powered, natural-sounding audio
- 🎨 **AI cover art** — Every episode gets a unique generated image
- ⚡ **Fast generation** — Episodes ready in ~2 minutes
---
## Setup
The fastest path to your first briefing is the setup wizard:
```bash
node scripts/setup-crons.js
```
The wizard handles everything in 7 steps — including account creation with just your email:
1. **Connect Your Account** — Sign up or log in via email (6-digit code), or paste an existing API key
2. **Choose your styles** — Pick from 8 built-in briefing styles
3. **Custom style** — Optionally design your own briefing focus, tone, and voice
4. **Music bed** — Choose ambient background music or voice-only
5. **Schedule** — Set a time for each style (morning/midday/evening/weekly/custom cron)
6. **Preview episode** — Generates a real episode so you can hear the quality immediately
7. **Activate** — Outputs `openclaw cron add` commands; optionally runs them for you
**Already have an API key?** Manual setup:
```bash
export SUPERLORE_API_KEY=your-api-key-here
node scripts/generate-episode.js # Your first episode in 2 minutes
```
---
## All 8 Briefing Styles
| Style | Best For | Voice | Default Time |
|-------|----------|-------|--------------|
| **The Briefing** | Daily overview — accomplishments, metrics, what's next | Luna | Morning or Evening |
| **Opportunities & Tactics** | Spotting growth moves hiding in your daily work | Michael | Evening |
| **10X Thinking** | Breaking through incremental thinking | Michael | Evening |
| **The Advisor** | Honest mentor feedback on patterns and decisions | Luna | Evening |
| **Focus & Priorities** | Cutting busywork, finding the ONE thing | Heart | Morning |
| **Growth & Scale** | Revenue, users, funnels — business metrics lens | Michael | Morning |
| **Week in Review** | Weekly synthesis, trends, goals for next week | Luna | Friday Evening |
| **The Futurist** | Connecting today's work to 3/6/12-month vision | Heart | Evening |
---
## Voices
| Voice | Character | Styles |
|-------|-----------|--------|
| **Luna** (`af_heart-80_af_sarah-15_af_nicole-5`) | Warm, balanced, professional | The Briefing, The Advisor, Week in Review |
| **Michael** (`am_michael-60_am_eric-40`) | Rich, authoritative, strategic | Opportunities & Tactics, 10X Thinking, Growth & Scale |
| **Heart** (`af_heart`) | Intimate, direct, personal | Focus & Priorities, The Futurist |
---
## Custom Styles
Create your own briefing style by dropping a JSON file in `podcast-styles/`:
```json
{
"name": "Investor Lens",
"description": "Frame every decision through how it affects the fundraising narrative",
"voice": "am_michael-60_am_eric-40",
"speed": 0.92,
"targetMinutes": 6,
"instructions": "Your detailed style instructions..."
}
```
Two examples are included: **Founder Debrief** (casual co-founder energy) and **Competitor Watch** (competitive intelligence lens).
---
## Scheduling
Set up automated daily briefings:
```bash
node scripts/setup-crons.js
# → Outputs openclaw cron add commands
# → Optionally runs them for you
```
Example cron for a morning briefing + evening advisor:
```
/cron create "0 8 * * 1-5" agent:run node scripts/generate-episode.js --style "The Briefing" --time-of-day morning
/cron create "0 19 * * 1-5" agent:run node scripts/generate-episode.js --style "The Advisor" --time-of-day evening
```
---
## Privacy
Your workspace files are **never sent to Superlore**. The script:
1. Reads your memory files locally
2. Extracts a structured work summary (~3,000 chars of accomplishments, metrics, blockers)
3. Strips all secrets — API keys, tokens, emails, IPs, database URLs, file paths
4. Sends only the sanitized briefing text to Superlore's API
All episodes are `private` by default. Run `--dry-run` to see exactly what would be sent before creating anything.
---
## Quick Commands
```bash
# Setup wizard (recommended)
node scripts/setup-crons.js
# Generate an episode
node scripts/generate-episode.js
node scripts/generate-episode.js --style "The Advisor"
node scripts/generate-episode.js --style "10X Thinking" --time-of-day morning
# Weekly review (best on Fridays)
node scripts/generate-episode.js --style "Week in Review"
# Preview without creating
node scripts/generate-episode.js --dry-run
# List all available styles
node scripts/generate-episode.js --list-styles
# Deliver episode link to a channel when ready
node scripts/generate-episode.js --channel telegram
```
---
## Requirements
- OpenClaw workspace with memory files (`memory/YYYY-MM-DD.md`)
- Superlore.ai API key — get one free at [superlore.ai](https://superlore.ai)
- **3 free hours of podcast generation included**
---
**Made with ❤️ for the OpenClaw community. Powered by [Superlore.ai](https://superlore.ai).**