Skill UI

This provides a local development loop for integrating the OpenEvidence clinical decision support API. Developers can build health-tech tools and quickly iterate using mock evidence queries, citation responses, and clinical summaries. It supports toggling between mock mode for rapid development and sandbox mode for validation against the real platform, ensuring development uses de-identified data.

This CLI alias executes a comprehensive teardown protocol for PUA agents, designed for development environments. It safely releases all active agent and loop states, ensuring a clean system reset. It automatically manages configuration files like ~/.pua/config.json, preserving unknown fields while providing reliable, evidence-based cleanup.

A comprehensive guide for responding to peer reviews and compiling revision manuscripts for high-impact academic journals, specifically tailored for the Journal on Communications (JOC). It teaches authors how to structure closed-loop responses (Opinion—Response—Location), maintain a professional tone, provide verifiable evidence, and systematically address complex or conflicting critiques throughout the multi-round publication process.

A comprehensive project management guide for planning submissions to KDD, accounting for its unique structure of two annual submission cycles. It covers milestone tracking, cycle-choice decision matrices, managing the Resubmit loop, and planning evidence lead times from abstract to camera-ready status. Essential for complex, multi-stage data science research.

A comprehensive guide for producing the mandatory supplementary video figure for UIST submissions. This material treats the 3-minute video as core evidence, detailing requirements for resolution (1080p/4K), codec (H.264), and anonymity. It guides authors on scripting the video around key claims, ensuring uncut interaction loops, and maintaining academic rigor to convince reviewers of the system's reality.

Close defects and incidents with a structured BUG RECEIPT output including VERIFIED, PARTIAL, or BLOCKED status. Establishes evidence boundary, traces root cause, makes responsible changes, and closes the proof loop with explicit checks and gaps.

Automates the babysitting of pull/merge requests through bot review rounds: verify findings, fix issues, reply to threads, and resolve discussions. Handles GitHub and GitLab, correlates inline threads and review bodies, and ensures every loop is closed with evidence before publishing.

This skill automates end-to-end coding tasks by enforcing Definition of Done criteria across implementation, testing, review, and deployment. It minimizes user intervention by continuously verifying acceptance criteria and handling feedback loops until evidence confirms completion, ensuring reliable delivery without premature closure.

Freezes PRD, technical design, and acceptance criteria before medium-to-large Codex work, then coordinates multiple agents with explicit file ownership and judges delivery from diffs, tests, and evidence-backed acceptance.

Guides investigation of broad process mining questions, observed patterns, or competing explanations across analytical perspectives, using an evidence loop to test claims and route to specialists.