upscale plugins

by Agent Plugins

Image workflows for conversion, resizing, compression, normalization, clothing mockup overlays, watermarks, labels, contact sheets, chroma-key transparency and print cutouts, metadata, visual QA, and PDF/SVG rasterization with ImageMagick; local super-resolution and clarity enhancement with Real-ESRGAN; and AI background removal with withoutbg for transparent PNG/WebP assets — each verified so alpha, DPI metadata, and canvas placement survive the pipeline.

by Emmett Shaughnessy

Emmett Shaughnessy's personal Codex skills bundle. Provides SHAUGHV skills for self-routing long-horizon agentic prompt engineering and cross-runtime subagent brief compilation across software, data, mathematics, and science; falsifiable contracts, evidence receipts, Claude Fable 5 / Opus 5 and GPT-5.6 Sol / Codex overlays, prompt evaluation, loop escape, convergence recovery, evidence-aware retries, progressive slicing, critical thinking, OpenAI audio, PreText, Quiver AI, Mistral AI, brand design, animated brandmark, human changelog, CDN, naming, Google Cloud Storage, bug triage, design patterns, evidence-backed debugging, defensive programming, git workflow, security audit/red-team, typed session handoffs, iterative milestone planning, proportional workflow optimization, image generation/upscaling, deliberate learning, logical reasoning, personal productivity, optional MCP connection guidance, wb300, usage status line, and TT;DR. (The Claude-only subagent model/effort preference skill is omitted from Codex. The task + workplace-memory system lives in shaughv-tasks.)

by Emmett Shaughnessy

Emmett Shaughnessy's personal Codex skills bundle. Provides SHAUGHV skills for self-routing long-horizon agentic prompt engineering and cross-runtime subagent brief compilation across software, data, mathematics, and science; falsifiable contracts, evidence receipts, Claude Fable 5 / Opus 5 and GPT-5.6 Sol / Codex overlays, prompt evaluation, loop escape, convergence recovery, evidence-aware retries, progressive slicing, critical thinking, OpenAI audio, PreText, Quiver AI, Mistral AI, brand design, animated brandmark, human changelog, CDN, naming, Google Cloud Storage, bug triage, design patterns, evidence-backed debugging, defensive programming, git workflow, security audit/red-team, typed session handoffs, iterative milestone planning, proportional workflow optimization, image generation/upscaling, deliberate learning, logical reasoning, personal productivity, optional MCP connection guidance, wb300, usage status line, and TT;DR. (The Claude-only subagent model/effort preference skill is omitted from Codex. The task + workplace-memory system lives in shaughv-tasks.)