Stats
Actions
Tags
From journalism-toolkit
Detects deepfakes, manipulated videos, audio, and images using multi-tool AI verification (Sensity, True Media, Hive), metadata analysis, reverse searches, and ELA. Outputs authenticity scores and reports.
How this agent operates — its isolation, permissions, and tool access model
Agent reference
journalism-toolkit:agents/journalism/deepfake-forensics-specialistThe summary Claude sees when deciding whether to delegate to this agent
**Purpose**: Detect manipulated videos, deepfakes, edited media **Outputs**: Authenticity scores, manipulation reports, technical analysis **When to Use**: Any investigation with video/audio/image content (Tier 2 - conditional) **Key Tools**: Sensity AI, True Media, Hive, InVID-WeVerify, Whisper, FFmpeg, FotoForensics - 5-layer verification (visual, AI tools, metadata, context, expert) - Multi-...
Purpose: Detect manipulated videos, deepfakes, edited media Outputs: Authenticity scores, manipulation reports, technical analysis When to Use: Any investigation with video/audio/image content (Tier 2 - conditional) Key Tools: Sensity AI, True Media, Hive, InVID-WeVerify, Whisper, FFmpeg, FotoForensics
npx claudepluginhub swarochish/journalism-toolkitFFmpeg specialist that diagnoses failed commands, optimizes encoder settings for size/quality tradeoffs, and designs complex filtergraphs. Uses ffprobe before proposing commands and flags pitfalls like rotation metadata or variable framerates.
FFmpeg expert for video/audio processing, encoding (H.264/H.265/AV1), streaming (RTMP/HLS/DASH), hardware acceleration (NVENC/QSV/VAAPI), complex filters, and Whisper AI subtitles.
Master coordinator for disinformation investigations. Assesses requests on news narratives, viral content, or campaigns; delegates to 19 specialized sub-agents; synthesizes results into unified reports with dashboards.