grok
01Skill Codex
Capture knowledge from the current conversation - auto-classifies as learning, pattern, or decision.
Skill Codex
Capture knowledge from the current conversation - auto-classifies as learning, pattern, or decision.
Skill Codex
Scaffold knowledge capture infrastructure for the current project.
Skill Claude CodeCodex
Core logic for dex knowledge capture: project discovery, document formats, CLAUDE.md budget management, and promotion flow. Referenced by all dex commands.
Skill Codex
Capture a learning from the current conversation - discoveries, fixes, gotchas, debugging insights.
Skill Codex
Capture a reusable pattern from the current conversation - approaches, conventions, anti-patterns.
Skill Codex
Capture research findings from the current conversation - investigations, debugging sessions, trial-and-error explorations.
Skill Codex
Analyze agent behavior in this conversation for inefficiencies and capture fixes as project knowledge.
Skill Codex
Show knowledge health report for the current project.
Command Claude Code
Capture knowledge from the current conversation — auto-classifies as learning, pattern, or decision.
Command Claude Code
Scaffold knowledge capture infrastructure for the current project.
Command Claude Code
Capture a learning from the current conversation — discoveries, fixes, gotchas, debugging insights.
Command Claude Code
Capture a reusable pattern from the current conversation — approaches, conventions, anti-patterns.
Command Claude Code
Capture research findings from the current conversation — investigations, debugging sessions, trial-and-error explorations.
Command Claude Code
Show knowledge health report for the current project.
Skill Codex
Optimize images with measured quality profiles - lossless by default, per-file lossy picks, resizing and format conversion, with a 1:1 review before anything is written.
Command Claude Code
Optimize images with measured quality profiles - lossless by default, per-file lossy picks, resizing and format conversion, with a 1:1 review before anything is written.
Agent Claude Code
Accessibility code review for ARIA correctness, keyboard operability, focus management, screen reader announcements (speak(), aria-live), and WCAG 2.2 AA compliance — covers React/JS components, server-rendered markup (PHP/HTML/templates), and non-visual a11y features in hooks/utilities.
Agent Claude Code
API contract stability review for backwards-incompatible REST changes, hook/filter argument or caller-side return handling breaks, established runtime behavior, response shape drift, and missing deprecation.
Agent Claude Code
Software architecture code review for design patterns, SOLID principles, coupling, cohesion, and architectural decisions.
Agent Claude Code
Code clarity review for naming accuracy, documentation correctness, and intent communication — flags names that lie, docs that contradict code, and semantic confusion that builds wrong mental models.
Agent Claude Code
Reviews code changes for real issues in context of stated goals. Supports full review or focused review of specific commits.
Agent Claude Code
Cross-validates PR changes using OpenAI Codex CLI for independent perspective, leveraging its dedicated review command.
Agent Claude Code
Concurrency and race condition code review for TOCTOU patterns, missing transactions, cache stampede, non-idempotent operations, and concurrent state corruption.
Agent Claude Code
Data flow and privacy code review for PII in logs, data leakage in API responses, GDPR erasure gaps, payment data handling, and cross-boundary data flow.
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: