Claude Skills by theheavenlyd3mon

Full-novel craft + anti-slop pipeline with project mode (ledger, characters, foreshadow, worldbuilding bible), draft→review→revise gates, and stability-trap enforcement at scene/chapter/manuscript scale. Single-scene mode retained. Use for fiction, lore, character docs, dialogue, worldbuilding, or book projects.

Author live widget apps for the Hermes TUI dock.

Use whenever a task involves building or editing a 3D scene or asset in Blender — modeling, characters/people, architecture/interiors, terrain/landscapes, props, materials, lighting, or rendering. Covers discovering installed add-ons, using generators, importing and REFINING ready-made assets, and matching the result to the spec. Requires a running Blender instance with the blender-mcp addon (see Prerequisite).

Vision-Language MCP tools for reading/visualizing any file (images, videos, documents, code, data, 3D models, etc.) and calling VL APIs (caption, ASR, search, etc.).

Generate step-by-step math problem-solving tutorial videos in Chinese (Mandarin). Use when: (1) a user provides a math problem and wants an explanation video, (2) someone says "make a math tutorial", "explain this equation", "create a teaching video for this problem", "讲解这道题", "生成解题视频", (3) the user wants a Chinese-language math lesson covering formulas, equations, or geometric figures, (4) the user shares a math problem in text or LaTeX and asks for a video walkthrough, (5) the input is an i...

Use whenever a task involves parametric CAD in FreeCAD — modeling parts and assemblies, editing object properties, technical drawings, importing/exporting STEP/STL/OBJ/DXF, PDF/Excel reports from a model, or finite-element (FEM/CalculiX) analysis. Requires a running FreeCAD instance with the FreeCADMCP addon (see Prerequisite).

Omni-model audio/video understanding MCP tools: segmented captioning with timestamps, ASR (plain / controllable-granularity / multi-speaker diarized), temporal grounding (locate a segment by text query), and event/action counting. Use for questions about WHAT is said/happens in an audio or video file and WHEN, on clips up to a few minutes.

Editing-director skill that OWNS every video task built from EXISTING REAL FOOTAGE the user supplies (vlog, montage, intro, recap, eating/travel/family edits, style replication, compositing, subtitles, voiceover, B-roll). When footage files are the input, use THIS skill first — not the generic hyperframes entry and not the general-video workflow: it contributes footage judgment (selection, pacing, beat-sync, sound, looks, per-scene design) and then hands the designed composition to the HyperF...

Triggered for long videos (30+ minutes), whether a single file or a directory of multiple videos. Vision-language MCP tools designed for efficient reading and semantic analysis of long videos (30+ minutes), supporting memory construction and semantic search.

Safe web scraping and research: detect and neutralize prompt injection, code injection, and content-based attacks in scraped web content before it enters the agent's context.

Use when a user drops a large document packet to review.

Skills system architecture — progressive disclosure, frontmatter, platform filtering, self-editing

Recover and review content from X/Twitter links — especially long-form X Articles (x.com/i/article/ID) and gated posts — when the normal web tools fail, and review promotional X articles for affiliate/funnel framing. Use whenever the user drops an x.com link, says "review this article on X", or asks to read an X post/thread that web_extract or Firecrawl can't reach.

Triage a batch of AI-agent / LLM GitHub repos (or link dumps) for personal relevance — collapse star-inflated lists into the few that actually apply to your stack and workflow. Use when the user pastes a list of GitHub links and asks "which are worth it" / "review these resources" / "look into all these".

Use when running the full idea workflow: capture a rough idea, expand it into design/UI/implementation docs, research similar products, and generate build-ready Markdown artifacts.

Use when turning a rough idea into a focused product/design Markdown doc through guided questions, without moving into implementation too early.

Remediate npm vulnerability advisories on a Node project or fork with minimal risk. Clears in-range advisories via non-force `npm audit fix`, isolates force-required bumps (out-of-range) behind a human decision, and verifies the test suite stays green — including the NODE_ENV=production dev-dependency trap that makes `npm test` fail with "vitest: command not found". Trigger when the user says "npm audit shows highs", "remediate vulnerabilities", "update deps on the fork", "fix npm audit", or ...

Audit a code surface (module, package, integration, CLI, MCP/tool layer) for over-engineering and missing capabilities using the user's "Ponytail lens" (minimalist / xkcd-Ponytail engineering ethos: YAGNI, delete dead code, reuse in-codebase, kill duplication). Trigger when the user says "audit X for over-engineering / missing capabilities", "review under Ponytail lens", "find bloat / gaps in <module>", or asks to slim down or gap-check an integration, importer set, CLI, or tool surface. Prod...

Execute plans via delegate_task subagents (2-stage review).

Compress skill/config/prompt files using Proteus-style DSL — token packing, semantic normalization, DSL encoding, structural compression, state-machine loops, arrow conditionals. Reduces 40-70% on behavioral/header sections while keeping operational instructions readable.

Minimize tool call count per task by assessing context sufficiency before reaching for tools. Cross-cutting behavioral discipline for all agent tasks.

Write implementation plans: bite-sized tasks, paths, code.

Use when working with Browserbase or the browse CLI.

Notes on fetching web content ordinary scrapers cannot reach — the xurl article API and the Browserbase dual-control browser session.

Use when bulk-pulling llms.txt knowledge bases to markdown.

Use when reading a Markdown corpus to write one digest file.

Design and run an AI-assisted long-form fiction pipeline that turns small story ideas into publishable short novels / audiobooks. Covers the book-writer profile architecture, specialist-profile delegation lanes, manuscript folder convention, and the draft→review→revise→export loop. Use when the user wants to mass-produce books via Hermes, build a book-writer agent, or automate novel/audiobook production.

Manage Ghost CMS content from the terminal — create and list posts, pages, and tags, and fetch site info via the Ghost Admin API (v5/v6). Use when the user asks about ghost, cms, blog, blogging, posts, pages, tags, publishing, or site configuration.

Edit drafts into sharper, more human writing while preserving the writer's personal voice, or detect AI-slop patterns without rewriting. Use when the user wants a draft clearer, more direct, more opinionated, or less AI-sounding, or asks whether writing reads as AI.

Author live widget apps for the Hermes TUI dock.

Review a content draft (X post, thread, LinkedIn post, blog, email) against brand voice and authenticity standards before it ships. Flags AI-isms, corporate jargon, unsubstantiated claims, manufactured stakes, and voice inconsistencies. Provides severity-rated findings with before/after fixes. Use after drafting and before posting.

Build-in-public X (Twitter) growth strategy using Hermes as an autonomous content engine — profile design, VPS deployment, content pipeline, engagement cadence, and X Premium tier guidance for indie hackers.

Triage a folder of AI assets into tiers before drafting.

Draft a single piece of social content — X post, X thread, LinkedIn post, blog post, or short email — with channel-specific structure, hook patterns, and brand-voice application. Use when the user asks to "write a tweet about X", "draft a thread on Y", "post about today's progress", or any other single-asset content request. Pairs with brand-review (check before posting) and build-in-public (strategy layer).

Research an X account or specific post before drafting a reply, then draft a value-add reply (not a pitch, not generic, not spam). The #1 growth lever for accounts under 5K followers — reply prioritization rewards quality engagement on bigger accounts. Use when targeting a specific tweet or account for engagement, when drafting reply threads, or when running the daily engagement cron.

Use when researching X trends or creating content calendars.