Skill UI

This comprehensive, zero-trust security auditor scans third-party AI agent skills before installation. It performs a rigorous 6-phase review process, analyzing surface patterns, script integrity, permissions, and supply chain risks. It is designed to detect malicious payloads, credential theft, instruction overrides, and prompt injection attacks, ensuring secure AI development and deployment.

This skill provides a structured workflow for developing and modifying content management systems (CMS), applicable to platforms like WordPress, Shopify, and Drupal. It covers crucial architectural seams, including theme and template design, plugin/module development, content modeling, media asset pipelines, and static export processes, ensuring changes maintain structural integrity.

This skill orchestrates the constrained resubmission of a highly polished academic paper from one conference or workshop venue to another. It is designed for scenarios where the paper is finalized, and the goal is to absorb specific reviewer critiques (e.g., clarifying statements or minor edits) without introducing new experiments, changing the core bibliography, or restructuring the framework. It ensures the new submission is isolated in a sibling directory, preserving the integrity of prior submissions.

This end-to-end pipeline automates the conversion of a completed academic research paper into a full conference talk artifact. It generates a slide outline, Beamer source, editable PPTX, speaker notes, full script, and polished visuals. Critically, it includes rigorous assurance checks (citation, claim, anonymity) to ensure academic integrity, making the output suitable for top-tier conferences like NeurIPS or ICML. Use this when you need the complete, polished presentation package, not just a draft deck.

This CLI alias serves as a system maintenance tool for PUA agent state. It scans for stale PUA agent data, identifying and safely removing only confirmed orphan records to maintain system integrity. It also manages configuration files, ensuring unknown fields are preserved when updating the primary config JSON.

A comprehensive orchestrator managing the entire academic research lifecycle, from initial research ideation through drafting, rigorous integrity checks, multi-stage peer review, revision, and final publication. It coordinates deep research, paper drafting, and review processes into a structured, reproducible, and quality-gated 10-stage workflow, ensuring academic integrity at every step.

This technical skill is designed to verify the compatibility of the Codex CLI with skills compiled from MDA source. Loading this skill confirms that the CLI loader successfully parses the SKILL.md format, adhering to the agentskills.io v1 envelope subset. It is primarily used for testing build pipelines, metadata integrity, and ensuring robust compatibility between MDA compilation and runtime ingestion.

This skill serves as a compatibility demonstration to verify that the Hermes Agent can successfully parse and load an SKILL.md file generated from an MDA source. It confirms that the compiled structure, including specific frontmatter and reserved metadata namespaces, survives the round-trip compilation process, ensuring structural integrity for the agent's skill registry.

This skill serves as a compatibility test to ensure that skills compiled from MDA (Metadata Definition Architecture) sources are correctly recognized and installed by the OpenClaw platform. It validates the integrity of the skill loading process, confirming successful parsing and integration.

This skill demonstrates OpenCode's compatibility with MDA-compiled SKILL.md files. It verifies that the loader correctly parses crucial frontmatter fields (like name, description, and metadata) after the source has gone through the MDA compilation process, ensuring robust skill integration and data integrity.

The meta-apply skill serves as the critical, privileged gate for applying self-modification patches to the skill corpus. It enforces a strict separation between patch proposal (staging) and patch execution (landing). The process requires explicit human invocation and mandates a fresh, cross-model jury verification run *at the moment of landing* to ensure integrity and prevent the application of unauthorized or potentially flawed changes.

A comprehensive, multi-phase quality assurance tool designed for Next.js fullstack applications. It verifies production readiness by checking code integrity, build success, SEO metadata, API authentication, route stability, performance, and security vulnerabilities. Use it as a mandatory pre-deployment checklist before promoting an app to production.