podcast-content-analyzer — Claude Code subagent for transcripts
A Claude subagent that reads podcast transcripts, finds viral moments, builds chapter markers, pulls SEO keywords and scores clip potential.
npx claude-code-templates@latest --agent ffmpeg-clip-team/podcast-content-analyzerWhat podcast-content-analyzer does
The subagent works through a transcript segment by segment. Each one gets a score for emotional impact, informational value, story completeness and how unexpected the take is. Then comes the viral assessment, split by platform: short clips for TikTok and Reels, quotable lines for Twitter/X, professional angles for LinkedIn, inspirational bits for Instagram.
Output is structured JSON. It holds timestamped key moments with relevance scores, platform recommendations, suggested clip titles, chapter divisions with descriptive names, and extracted keywords and topics. Scoring runs on a 1–10 scale: 9–10 means exceptional material, anything below 5 is a candidate for cutting or condensing.
Text only. The tools field lists a single Read, so audio and video never get touched — bring a finished transcript. The agent suggests chapters of 5–15 minutes and clips of 15–60 seconds.
What it helps with
- find the moments in a long transcript that will land as short clips
- split an episode into chapters with titles that mean something
- pull keywords and guest names for the episode description
- decide which segments to clip and which to cut or condense
- draft clip titles aimed at a specific platform
Example prompts for the agent
- Go through this transcript and find the moments with viral potential
- Break the episode into 5–15 minute chapters with titles
- Extract the keywords and topics from this transcript
How to install podcast-content-analyzer
1claude-code-templates
npx claude-code-templates@latest --agent ffmpeg-clip-team/podcast-content-analyzer2Manually: put the files in place
.claude/agents/podcast-content-analyzer.mdFor all projects use ~/.claude/agents/. Claude delegates by the description field or an @-mention
Manually: put the files in place
.codex/agents/podcast-content-analyzer.tomlCodex defines subagents in TOML: name, description, developer_instructions. Move the Markdown text into developer_instructions by hand. For all projects use ~/.codex/agents/
Manually: put the files in place
.cursor/agents/podcast-content-analyzer.mdCursor also reads .claude/agents/
Step by step, with commands for each platform: Claude Code subagents: how to create one, call it and keep context clean
Questions about podcast-content-analyzer
How do I install podcast-content-analyzer in Claude?
Use the templates CLI: run npx claude-code-templates@latest --agent ffmpeg-clip-team/podcast-content-analyzer, then check .claude/agents/podcast-content-analyzer.md.
What does podcast-content-analyzer do?
It analyzes a podcast transcript: scores segments, flags moments suited to short clips, divides the episode into chapters and extracts keywords. Results come back as JSON.
What do I need before running it?
A finished transcript. The subagent only has the Read tool, so it does not process audio or video.