Reusable agent skills for practical writing, workflow packaging, Codex goal drafting, and Codex app run actions.
This repository also includes a sanitized root AGENTS.md derived from my global Codex instructions, with private registry and cluster details replaced by placeholders.
logseq-writer- turns topics, drafts, and notes into practical Logseq-style tutorial articles.workflow-packager- reviews recent work evidence and recommends minimal reusable workflow assets.codex-goal-builder- turns rough objectives into evidence-based Codex Goals.codex-runner-creator- creates or repairs.codex/environments/environment.tomlrun actions.git-commit-push- creates session-scoped Conventional Commits and safely pushes the new commit(s) to the tracked remote branch.pr-creator- creates pull requests with template compliance and explicit fork/upstream head safety.screenshot-interaction- infers UI behavior and missing states from screenshots before implementation.issue-creator- turns terse QA notes into structured GitHub issues with gh preflight, safe metadata, and sensitive-info checks.
Install all skills:
npx skills add mlhiter/skills --allInstall one skill:
npx skills add mlhiter/skills --skill logseq-writerPreview the repository first:
npx skills add mlhiter/skills --listSkills are published under the canonical skills/<skill-name>/SKILL.md structure used by skills.sh. The root AGENTS.md documents the shared agent operating rules for this repository:
AGENTS.md
skills/
logseq-writer/SKILL.md
workflow-packager/SKILL.md
codex-goal-builder/SKILL.md
codex-runner-creator/SKILL.md
git-commit-push/SKILL.md
pr-creator/SKILL.md
screenshot-interaction/SKILL.md
issue-creator/SKILL.md
These skills are instruction bundles. Read each SKILL.md before installing, especially skills that inspect local histories, write project files, or create automation assets.