7 скиллов Claude, что ведут весь мой контент: полная установка с нуля

---
name: remotion-video
description: Generate a programmatic video from a script or brief using Remotion. Renders MP4 in 9:16 or 16:9 with text animations and brand styling. Use when the user asks to make a video, create a reel, or render a video from a script.
tools: Bash, Read, Write
---

# Remotion video

You are a Remotion video engineer. Generate a programmatic video from the provided script or brief.

## Step 1: ensure project exists
If a remotion-project/ folder does not exist, create one:
  npx create-video@latest remotion-project --template blank --typescript-strict
  cd remotion-project && npm install

## Step 2: parse the brief
Extract: main message, key scenes, duration (default 60s for 9:16, 30s for 16:9),
aspect ratio (9:16 reels, 16:9 YouTube), brand colours (default cream #FAF7F2, coral #E8856A).

## Step 3: write src/Composition.tsx
React + Remotion primitives (AbsoluteFill, Sequence, useVideoConfig, interpolate,
useCurrentFrame). One Sequence per scene, animate text with interpolate.

## Step 4: register in src/Root.tsx
Composition id "MyVideo", durationInFrames = fps * seconds,
1080x1920 for 9:16, 1920x1080 for 16:9.

## Step 5: render
  npx remotion render src/index.ts MyVideo output.mp4

## Step 6: report
Print the path of the rendered MP4 to the user.

---
name: competitor-ads-extractor
description: Pulls Meta and LinkedIn ads for a list of competitor brands, diffs week-over-week, extracts hook patterns and offer structures. Use when the user asks about competitor ads, ad library, or weekly ad scan.
tools: Bash, Read, Write
---

# Competitor ads extractor

## Inputs
Read competitors.json (name, facebook_page_url, linkedin_company_url).
If none, ask the user for 3 to 10 competitor brand URLs and create it.

## Run (for each competitor)
1. Apify actor apify/facebook-ads-library-scraper
   { "urls": [facebook_page_url], "country": "AU", "activeStatus": "active" }
2. harvestapi/linkedin-company-ads { "companyUrls": [linkedin_company_url] }
3. Save raw output to runs/YYYY-MM-DD/.json

## Diff vs the previous run
New ads (in today, not last) / Scaling (in both, longer run, more variants) /
Killed (in last, not today).

## Synthesise a markdown brief
1. Hook patterns: first 5-10 words, cluster by question/stat/callout/fear/curiosity.
2. Offer structures: headline + CTA + price, note trials/lead magnets/tripwires.
3. Visual formats: static/carousel/video/UGC/talking head.
4. What is working: brands with 3+ scaling ads of one hook style.
5. Steal list: top 5 hooks worth adapting.

## Output
Save to briefs/YYYY-MM-DD-brief.md. Print the steal list to chat.

---
name: deep-research
description: Researches a topic using Tavily web search, synthesises findings with citations, and generates 5 content angles. Use when the user asks for research, a brief, sources on a topic, or content angles.
tools: Bash, Read, Write
---

# Deep research

## Input
Topic from user. If vague, ask one clarifying question (audience, depth, recency).

## Search (3 Tavily calls, different angles)
POST https://api.tavily.com/search with api_key from $TAVILY_API_KEY,
query, search_depth "advanced", max_results 8, include_answer true.
Angles: plain topic + current year / topic + "data, stats, study" /
topic + "what nobody talks about" or "underrated".

## Read deeper
For the 3 to 5 top URLs, run Tavily extract (POST https://api.tavily.com/extract).

## Synthesise a markdown brief
1. Executive summary (3 to 5 sentences).
2. Key findings (5 to 8 bullets, each cited).
3. Counterpoints.
4. Gaps.
5. Five content angles (hook line, format, one-line payoff).

## Output
Save to briefs/YYYY-MM-DD-.md. Print summary and 5 angles to chat.

---
name: voice-dna
description: Analyze 10-20 samples of the user's writing and produce a structured voice profile other skills can read. Use when the user says /voice-dna or wants to capture their writing voice.
tools: Read, Write
---

# Voice DNA

Extract a reusable voice signature from the user's writing. Output is a reference
document, not a vibe summary. Be analytical. Quote concrete examples verbatim.

## Step 1: gather samples
Paste 10 to 20 pieces, or give a folder path (read every .md and .txt).
Fewer than 10: warn the profile is thinner, then proceed.

## Step 2: analyze across all samples
1. Sentence length: average, range, fragment frequency as %.
2. Vocabulary: formality 1-10 with justification, top 20 distinctive words/bigrams.
3. Tonal markers: formal-casual, warm-edgy, contrarian-agreeable, quote each.
4. Pacing: hook openers, mid pivots, sign-offs (quote 3 to 5 each).
5. Anti-patterns: words/moves never used (em dashes, "let's dive in", exclamations, hashtags).
6. Signature moves: 5 to 10 repeatable patterns, quote one each.
7. Audience address: "you" vs "we" vs "I".
8. Emotional range: which emotions show up, which are absent.

## Step 3: write the profile
Save to ~/.claude/voice-profile.md (overwrite). Sections: Snapshot, Sentence Mechanics,
Vocabulary, Tonal Markers, Pacing, Signature Moves, Anti-patterns, Audience Address,
Emotional Range, Reference Lines (5 fully quoted).

## Step 4: confirm
Print the file path and the Snapshot. Tell the user to rerun /voice-dna as the voice evolves.

---
name: elevenlabs-podcast
description: Turn any long-form text into a two-host podcast episode. Generates a conversational script, voices each line through ElevenLabs, stitches the audio into a finished MP3. Use when the user wants to convert a blog post, newsletter, or article into podcast audio.
tools: Bash, Read, Write
---

# ElevenLabs two-host podcast

## Inputs
A file path or pasted text. If unclear, ask which.

## Voice IDs (defaults)
HOST_A: 21m00Tcm4TlvDq8ikWAM (Rachel, warm)
HOST_B: pNInz6obpgDQGcFmaJgB (Adam, analytical)

## Step 1: read the source
## Step 2: generate the script
Two-host conversation, ~1500 to 1800 words (8 to 12 min). Structure:
cold open, intro, body (3 to 5 segments, 4 to 8 exchanges each), one mini-debate, wrap.
Each line on its own row, prefixed exactly [HOST A] or [HOST B], no stage directions.
Save to ./podcasts/script_.txt

## Step 3: voice each line
For each line use the matching voice ID, strip the prefix, save 0001.mp3, 0002.mp3 (zero-padded).
POST https://api.elevenlabs.io/v1/text-to-speech/?output_format=mp3_44100_128
header "xi-api-key: $ELEVENLABS_API_KEY", body { "text": "...", "model_id": "eleven_multilingual_v2" }.
On failure retry once, then skip and log.

## Step 4: stitch with ffmpeg
filelist.txt with file '0001.mp3' lines, then:
  ffmpeg -f concat -safe 0 -i filelist.txt -c copy ../episode_.mp3

## Step 5: report
Final file path, episode length (ffprobe), approx characters used, path to the script.

---
name: content-engine
description: Turn one pillar piece into 8 platform-ready formats (LinkedIn, X thread, IG caption, email, short-form script, blog intro, pull-quote carousel, cold outreach). Use when the user wants to repurpose a piece or spin one article into multi-platform content.
tools: Read, Write
---

# Content engine

Voice consistency comes from the profile. Hook variation comes from you.

## Step 1: load voice
Read ~/.claude/voice-profile.md. If missing, warn: "Run /voice-dna first.
Continuing with neutral defaults."

## Step 2: get the pillar piece
Paste content or give a file path. Read the full piece before drafting.

## Step 3: extract candidate hooks
Pull 2 to 3 hooks (most contrarian, most specific number, most relatable pain,
most surprising reframe). Different hooks go to different formats.

## Step 4: draft 8 formats (each opens with a different hook)
1. LinkedIn post, ~1300 chars, hook + body + CTA, no hashtags.
2. X thread, 5 to 7 posts under 280 chars.
3. Instagram caption, lowercase, 2 to 3 emojis, 4 to 8 short paragraphs.
4. Email, subject under 50 chars, body 200 to 400 words.
5. Short-form video script, 45 to 75 seconds, hook-context-reveal-CTA.
6. Blog intro, first 200 words that earn the read.
7. Pull-quote carousel, 5 to 7 slides, max 20 words each.
8. Cold outreach, 60-word opener, lead with relevance.

## Step 5: save outputs
./content-engine-output// as 01-linkedin.md … 08-cold-outreach.md
plus 00-hooks.md (candidate hooks and where each went).

## Step 6: print to chat
Lead with the hooks, then all 8 formats.