MCP App StoreMCP App Store
from ppt-image-first1,006
Turn vague presentation requests into image-first slide plans: clarify goals, propose visual directions, generate preview images, and produce deck specs before
PPT Image First converts a user's high-level presentation request into a generation-ready, image-first deck pipeline. The skill leads an agent through intake, content synthesis, style previews, and locked planning files (design_spec, slide_blueprint, spec_lock) before any bulk generation or PPT export. It enforces confirmation gates and preview-first workflows so visual direction is approved prior to final output.
Use when the user asks for a PPT, deck, slides, pitch, or similar—especially when they have a topic, rough notes, or brand constraints but need help with structure and visual direction. Triggers include 'make a presentation', 'create slides', 'deck for product demo', '答辩稿', and '路演 PPT'.
assets/preview_shell and other UI shells when producing review pages.Suited to agents that can drive image generation and file workflows (e.g., Codex/Claude Code, Cursor, or CLI-driven agents) and environments that can open local preview UIs.
PPT Image First is a presentation-planning skill that guides agents through a multi-phase workflow: intake, content basis, style preview with generated images, planning lock, and generation review. The sole script (render_review_markup.py) is a Pillow-based utility for overlaying review markup annotations onto slide images; it exited with an argument error in DRY_RUN mode since it requires a payload file and --images directory, which is expected behavior for a processing tool. The SKILL.md is well-structured with clear phases, confirmation gates, progressive loading, and hard rules — but it references multiple external files (references/, templates/, assets/) that were not bundled with the skill, making it partially incomplete without the full repo.
Clean skill — no security concerns. Script only does local image processing with Pillow, no network calls or suspicious patterns. Well-structured SKILL.md with clear I/O contract, progressive loading pattern, and hard rules. Main weakness is heavy dependency on unbundled reference/template/asset files that would need to be fetched from the full repo for the skill to be fully functional. The script is a utility rather than core logic, and its argument-error exit in DRY_RUN is expected.
Maestro Orchestrate
Orchestrates complex engineering workflows with enforced design and approval gates, child-agent delegation, and session state management.
Run Execute
Orchestrates execution of work items across modes (autopilot, confirm, validate) with scripted init/complete tooling, plan/test/report artifacts and strict gati
Colorize (Strategic UI Color)
Introduce purposeful color into monochromatic or dull interfaces to improve hierarchy, meaning, and emotional warmth while preserving accessibility.
Ip2Location IO Automation (Composio)
Automate Ip2location IO tasks via Composio's Rube MCP toolkit — discover tools, verify connections, and execute schema-compliant workflows safely.
Shopify Store Audit
Structured, page-by-page audit of Shopify storefronts to find conversion blockers and prioritized fixes for copy, UX, trust signals, and checkout friction.
PPT as Code
Creator-first workflow to plan and build HTML-first presentations with staged modes (quick/basic/advanced), optional PPTX export, and clear QA and image plannin
Build Unplanned Feature
Orchestrates ad-hoc feature development: requirement gathering, planning, implementation and validation across task-planning, implementation, and validation loo
Absolute Prune
Dead code and dependency cleanup across the entire repository.