xju2

@xju2 source repo

5 skills
0 followers
6 hours ago last updated

▌
Red Team Review · xju2
Produce a high-quality, constructive red-team review of a research proposal (DOE, NSF, NIH, or any competitive grant). Use this skill when the user asks to "red-team", "review", "critique", or "give feedback on" a proposal, grant application, or research plan, or when they want a reviewer's-eye view of how a proposal will be received. The review should be the kind of rigorous, expert, constructive critique that a strong panel reviewer would write — identifying real weaknesses, suggesting concrete fixes, and foregrounding genuine strengths.

0 installs
▌
Scientific Paper Writing · xju2 bundle
Draft, revise, or review scientific papers for clear structure, concise prose, consistent terminology and notation, defensible claims, and publication-ready figures, tables, equations, and references. Use for abstracts, titles, introductions, methods, results, discussions, conclusions, captions, full manuscripts, and paper-style technical writing, including AI/ML conference submissions. Follow the target venue and the user's explicit requirements over this skill's defaults.

0 installs
▌
Review Before Coding · xju2 bundle
Give brief, evidence-backed feedback on a requested software change before implementation. Use when the user asks to build, add, modify, refactor, migrate, optimize, remove, or otherwise change code, configuration, tests, infrastructure, or repository contents. Inspect only the project evidence needed to judge the request, answer whether it is worth doing, give concise feedback, then stop and wait for explicit user confirmation before making any coding changes. If the user is approving a review already completed by this skill and the scope is unchanged, proceed without repeating the review.

0 installs
▌
Harness Engineering Prompts · xju2 bundle
Create or revise prompts for AI coding agents such as Claude Code, Codex, or other software-engineering agents using OpenAI-style Harness Engineering principles. Use when the user asks for an implementation prompt, milestone prompt, issue-to-agent prompt, coding-agent task brief, or continuation prompt for an AI-driven software project. Ground the prompt in the target repository's own AGENTS.md, architecture docs, plans, ADRs, tests, and current code; emphasize outcomes and mechanically enforced invariants while preserving local agent autonomy, agent legibility, feedback loops, self-review, and improvements to the development harness.

0 installs
▌
Agent First Project Bootstrap · xju2 bundle
Bootstrap a new AI-driven or agent-first software project using Harness Engineering principles. Use when ChatGPT is asked to create, initialize, scaffold, or plan a new repository intended to be developed substantially by coding agents such as Claude Code, Codex, or similar systems. Establish repository-local knowledge, concise AGENTS.md guidance, architecture and milestone plans, deterministic setup and validation, executable acceptance criteria, mechanically enforced invariants, agent-legible observability, ADR and execution-plan conventions, and feedback loops that improve the development harness over time.

0 installs

5 published skills