kendreaditya/clean
claude/skills/clean/SKILL.md
Clean noisy machine-generated text into readable output. Unlike converting (pandoc), cleaning handles fragmented, duplicated, or metadata-heavy formats. Supports VTT and SRT subtitle files via cleansubs. Triggers: 'clean vtt', 'clean srt', 'clean subtitles', 'vtt to text', 'srt to text', 'vtt to markdown', 'clean transcript', 'subtitle to text', 'cleansubs'.
development
Updated Apr 17, 2026
$ install --global
skillsauthnpx skillsauth add kendreaditya/.config cleanInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 24, 2026, 9:03 PM1.9s1 file scanned
SKILL.md
- name:
- clean
- description:
- Clean noisy machine-generated text into readable output. Unlike converting (pandoc), cleaning handles fragmented, duplicated, or metadata-heavy formats. Supports VTT and SRT subtitle files via cleansubs. Triggers: 'clean vtt', 'clean srt', 'clean subtitles', 'vtt to text', 'srt to text', 'vtt to markdown', 'clean transcript', 'subtitle to text', 'cleansubs'.
Clean
Clean noisy, machine-generated text into readable output.
Converting (pandoc) = structured data, just change syntax. Cleaning = noisy data, strip metadata, dedup, stitch lines, find paragraph boundaries.
VTT/SRT Subtitles → Text/Markdown
CLI: cleansubs (at scripts/cleansubs.py)
# Plain text
cleansubs input.vtt # stdout
cleansubs input.srt # SRT directly, no conversion
cleansubs input.vtt -o output.txt # file
# Markdown with metadata (matches yt-research output format)
cleansubs input.vtt --md # stdout
cleansubs input.srt --md --title "Movie" -o t.md # SRT to markdown
cleansubs input.vtt --md --title "T" --id "abc" # with metadata
# Stdin
cat /tmp/VIDEO_ID.en.vtt | cleansubs
cat /tmp/VIDEO_ID.en.vtt | cleansubs --md --title "My Video"
# Batch — multiple files to one markdown
cleansubs /tmp/*.vtt --md --heading "Research" -o combined.md
# Force format when auto-detect is wrong
cleansubs input.vtt --format srt
cleansubs input.vtt --format youtube
Flags:
| Flag | Description |
|------|-------------|
| --md | Output markdown with ## Title, Video ID, URL, transcript, --- |
| --title | Video title (inferred from filename if omitted) |
| --id | Video ID (inferred from yt-dlp filename pattern ID.lang.vtt) |
| --url | Video URL (auto-generated from ID if omitted) |
| --heading | H1 heading for combined batch output |
| --format | Force format: auto (default), youtube, srt, structured_vtt |
| -o | Output file (stdout if omitted) |
Format auto-detection:
.srtextension or comma-timestamps → SRT pipeline (block-based, timing-gap paragraphs)Kind: captionsheader or no cue IDs → YouTube auto-gen VTT (overlap-merge dedup)- VTT with cue IDs → structured VTT (uses SRT-style pipeline)
What it cleans:
- Strips VTT/SRT headers, timestamps, cue IDs, Kind/Language metadata
- Strips HTML/VTT tags (timestamp spans, color tags, speaker tags)
- YouTube VTT: deduplicates overlapping fragments (substring + overlap merge)
- SRT/structured: block-based parsing, preserves commas and punctuation, timing-gap paragraphs (>2s gap = new paragraph)
- Keeps speaker labels and sound effects intact
- Fixes missing punctuation spacing
Use with youtube skill:
# Download subs + clean to plain text
yt-dlp --write-auto-sub --sub-lang en --sub-format vtt --skip-download \
-o "/tmp/%(id)s.%(ext)s" "URL" && cleansubs /tmp/VIDEO_ID.en.vtt
# Download subs + clean to markdown with metadata
yt-dlp --write-auto-sub --sub-lang en --sub-format vtt --skip-download \
-o "/tmp/%(id)s.%(ext)s" "URL" && \
cleansubs /tmp/VIDEO_ID.en.vtt --md --title "Video Title" -o transcript.md
Dependencies: Python 3 (stdlib only — no pip packages)
Related Skills
kendreaditya/test-coverage-advisor
testing
VerifiedTrustedCommunity
Reviews test coverage and suggests missing test cases for error paths, edge cases, and business logic. Activates when users write tests or implement new features.
SKILL.mdUpdated Jun 12, 2026
kendreaditya/tech-debt
development
VerifiedTrustedCommunity
Identify, categorize, and prioritize technical debt. Trigger with "tech debt", "technical debt audit", "what should we refactor", "code health", or when the user asks about code quality, refactoring priorities, or maintenance backlog.
SKILL.mdUpdated Jun 12, 2026
kendreaditya/security-audit
tools
VerifiedTrustedCommunity
Comprehensive security scanning and vulnerability detection. Includes input validation, path traversal prevention, CVE detection, and secure coding pattern enforcement. Use when: authentication implementation, authorization logic, payment processing, user data handling, API endpoint creation, file upload handling, database queries, external API integration. Skip when: read-only operations on public data, internal development tooling, static documentation, styling changes.
SKILL.mdUpdated Jun 12, 2026
kendreaditya/performance-optimization
development
VerifiedTrustedCommunity
Optimizes application performance. Use when performance requirements exist, when you suspect performance regressions, or when Core Web Vitals or load times need improvement. Use when profiling reveals bottlenecks that need fixing.
SKILL.mdUpdated Jun 12, 2026