Skill Codex
Implement specs with parallel agents, disjoint ownership, integration, and verification.
Skill Codex
Implement specs with parallel agents, disjoint ownership, integration, and verification.
Skill Codex
Create or check concise reviewer-visible PR context and practical proof when an authorized workflow publishes an update or prepares it for merge. Lead with what broke and how it is fixed, use actual product screenshots for static UI changes and recordings for motion or manual interaction, keep technical diagrams as…
Skill Codex
Review a GitHub PR in VS Code with local diffs, inline comments, navigation, and its worktree preserved.
Skill Codex
Audit PR diffs for unrelated artifacts, noisy comments, deletions, generated drift, and stray refactors.
Skill Codex
Review dense, clever, stringly typed, weakly typed, over-abstracted, or hard-to-read code.
Skill Codex
Investigate technical or product questions against primary sources, preserve substantive discoveries in the project Obsidian notes, and return cited conclusions.
Skill Codex
Map changed runtime flows through contracts and side effects to risks and proving validation.
Skill Codex
Manage budgets, scope, findings, consults, provisional fixes, and fixed-point rules for autonomous reviews.
Skill Codex
Run authorized native reviews and fixes until two fresh passes are clean.
Skill Codex
Find prior local Codex or Claude sessions; keep contents local and inspect cards before opening logs.
Skill Codex
Audit agent skills: loaded roots, duplicate skills, unused skills, prompt-budget costs, compact descriptions.
Skill Codex
Run before every final response: make it concise and clear, cut AI tells, re-pitch confusion, and use visuals only when they help.
Skill Codex
Build or fix behavior test-first through a red-green-refactor loop when the user requests TDD, test-first work, or integration-style tests.
Skill Codex
Plan and audit test portfolios before creating, changing, or removing tests or test infrastructure; challenge missing, duplicate, brittle, wrongly placed, ownerless, and dangerous coverage.
Skill Codex
Turn a resolved conversation into an Obsidian spec with testing seams and PR delivery shape.
Skill Codex
Turn a plan or spec into tracer-bullet Obsidian tickets with blockers and logical PR groups.
Skill Codex
Write or review TypeScript with shared types, boundary validation, safe narrowing, and verification.
Skill Codex
Wait for a command, CI run, subagent, or timed delay by holding one long call instead of polling; report only meaningful state changes.
Skill Codex
Write agent-facing skills, AGENTS.md, CLAUDE.md, and linked docs with precise behavioral instructions.
Skill Codex
Evidence-backed user corrections or self-detected mistakes revealing reusable agent failures: delegate one systemic-fix recommendation, get approval, then implement it.
Skill Claude CodeCodex
Select model-specific writing guidance for another skill, including exact coverage checks, same-family fallback, and one task-wide stale-coverage notice.
Skill Codex
variants/gpt-6.md.
Skill Codex
variants/gpt-6.md.
Skill Codex
variants/gpt-6.md.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: