Claude Commands
https://github.com/thedotmack/claude-commands- Category
- AI Agents
- Rank
- No. 1822Tools index
Previous survey · No. 1830 ·
- Pricing
- Open Source
- Type
- TOOL
- Builder
- thedotmack
- GitHub
- 11 stars
- Date
About
Custom Claude Code commands for ~/.claude/commands/.
What it does
A prompt-template kit that gives Claude Code specialized roles for planning, task execution, memory, logic analysis, and design audits. You insert a specific request into a shared placeholder, then run the chosen template through Claude Code. Several templates coordinate specialist agents or produce structured review artifacts.
Why it's ranked here
The collection is useful when repeated coding work benefits from firm, reusable instructions instead of fresh prompting. Its planning, orchestration, and design-audit templates are unusually detailed. The value is less certain outside the author’s setup because several workflows assume named specialist agents, while memory commands require separate claude-mem tools.
What's good
Each template defines a narrow job, concrete rules, and expected outputs. Planning stays focused on scope. Logic review separates analysis from implementation and records decisions for later action. The visual and technical design audits provide checkpoints, severity levels, evidence requirements, and report structures, making their results easier to compare and act upon.
Tradeoffs
The templates encode strong opinions, especially mandatory agent delegation and a make-it-work-first stance that rejects preemptive validation. That may conflict with security-sensitive or reliability-critical development. Memory depends on claude-mem and its storage tools. Design audits prescribe rigid thresholds and artifact structures, while some specialist agent names may not exist in another Claude Code setup.
How to use it well
It suits Claude Code users who repeatedly plan projects, delegate independent tasks, preserve session context, or run structured code and interface audits. Start with a precise request and adapt agent names to your environment. Treat the anti-defensive philosophy as a prototype heuristic, not a universal engineering rule. It does not supply an application, library, API, or general automation framework.
Technical notes+
README.md documents seven Markdown command templates and placeholder substitution through #$ARGUMENTS. think.md and do.md instruct Claude Code to spawn parallel agents; logic-blocks.md adds an interactive analysis protocol and progress artifacts. save.md and remember.md depend on claude-mem, while MANIFESTO.md documents its Chroma MCP retrieval tools and CLI storage command. design-review-visual.md specifies browser inspection, screenshots, breakpoints, and report output. design-review-technical.md specifies token, component, styling, accessibility, performance, and technical-debt audits. CLAUDE.md applies repository-wide agent and coding rules. No package manifest, source module, license file, or test directory appears among the supplied files.
Observed
- Primary format
- Markdown prompt templates
- Interface
- Claude Code CLI
- Install surface
- Template files used through placeholder substitution; no package manifest is shown
- Optional integration
- Memory commands require claude-mem CLI and Chroma MCP tools
- Command coverage
- Seven templates cover planning, execution, memory, logic review, and design review
- Repository structure
- No test directory appears among the supplied files
Read from README.md, do.md, save.md, think.md, CLAUDE.md, remember.md, MANIFESTO.md, logic-blocks.md, design-review-visual.md, design-review-technical.md.
What it can do
Execute custom Claude commands from command line
Command name and arguments → Command execution results
Store and organize reusable Claude code commands
Claude code scripts → Organized command library in ~/.claude/commands/
Create custom automation workflows with Claude
Workflow specifications and parameters → Automated task execution
Manage Claude command configurations
Configuration settings and command definitions → Updated command configurations
Run batch operations using predefined Claude commands
Batch job parameters and data → Processed batch results
Tags
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.