5 production-grade workflow skills for Claude Code.
Debugging, delegation, planning, code review, and configuration architecture - written from production experience and informed by the best open-source patterns.
Claude Code is powerful. But without workflow discipline:
- Debugging = random guessing until something works
- Delegation = manual, inconsistent, context-wasting
- Planning = "just start coding" then rewrite 3 times
- Code review = skipped or superficial
- CLAUDE.md = unmanageable blob at 20+ rules
5 skills that install in 30 seconds and immediately improve how you work:
| Skill | What Changes |
|---|---|
| Debugging | 4-phase cycle replaces random guessing. ACH method for complex bugs. |
| Delegation | Score-based routing to the right agent at the right model tier. |
| Plan & Execute | TDD plans that survive reality, with review gates between batches. |
| Code Review | Structured PR review with severity categorization and security focus. |
| Config Architect | Tiered CLAUDE.md that scales without wasting tokens. |
Each skill includes reference files with templates, checklists, and examples.
git clone https://github.com/primeline-ai/primeline-skills ~/.claude-plugins/primeline-skillsAdd to your Claude Code settings (~/.claude/settings.json):
{
"pluginDirectories": [
"~/.claude-plugins/primeline-skills"
]
}Restart Claude Code. All 5 skills are now available automatically.
If you only want one skill, copy its folder into your project's .claude/skills/ directory:
# Example: just the debugging skill
git clone https://github.com/primeline-ai/primeline-skills /tmp/primeline-skills
cp -r /tmp/primeline-skills/skills/debugging .claude/skills/Start Claude Code and ask it to debug something, plan a feature, or review code. It should announce "Using [skill-name]..." when the skill activates.
Every skill follows the same structure:
skills/[name]/
SKILL.md # The skill itself (Quick Start + Full System)
references/
[template].md # Ready-to-use templates
[guide].md # Deep-dive reference material
| Skill | References |
|---|---|
| Debugging | Evidence collection template, anti-patterns guide |
| Delegation | 5 prompt templates, model selection matrix |
| Plan & Execute | Plan template (full + minimal), review checklist |
| Code Review | Severity classification guide, security checklist |
| Config Architect | Copy-paste CLAUDE.md starter template, token measurement guide |
These skills are part of a progression. Each tier works independently - no hard dependencies.
You're here You want this Install this
----------- ------------- ------------
Raw Claude Code -> Session memory -> Starter System (free)
-> Workflow skills -> + Skills Bundle (free) <- you are here
-> Deep planning -> + UPF (free)
-> Deep analysis -> + Quantum Lens (free)
-> AI-powered system -> + Course (paid)
| Component | What It Does | Link |
|---|---|---|
| Starter System | Session memory, handoffs, context awareness | GitHub |
| Skills Bundle | 5 workflow skills (this repo) | You're reading it |
| UPF | Universal Planning Framework with 4-stage deep planning | GitHub |
| Quantum Lens | Multi-perspective analysis + solution engineering (7 cognitive lenses) | GitHub |
| Course | Kairn + Synapse: AI-powered memory and knowledge graphs | primeline.cc |
Skills reference each other with soft "if available" language:
- Debugging mentions
/rememberfor logging known failures (Starter System) - Plan & Execute mentions routing subtasks via the delegation skill (this bundle)
- All skills mention Kairn upgrade paths for AI-powered memory (Course)
If the referenced component isn't installed, nothing breaks. The tip is just skipped.
Each skill documents a "With Kairn" section explaining what becomes possible with AI-powered memory:
| Skill | Without Kairn | With Kairn |
|---|---|---|
| Debugging | Each session starts fresh | Pattern recognition across past bugs |
| Delegation | Manual scoring every time | Adaptive scoring from outcomes |
| Planning | Plans from scratch | AI-reviewed against past patterns |
| Code Review | Reviewer's memory only | Cross-project review patterns |
| Config | Manual organization | Auto-detection of needed rules |
The free skills work well on their own. Kairn makes them learn and improve over time.
- Quick Start for every skill. Beginners get value in 60 seconds.
- Full System for depth. Advanced users get the complete methodology.
- Zero dependencies. Pure markdown. No Python. No npm. No build step.
- No lock-in. MIT licensed. Copy, modify, redistribute.
- Security guardrails. Every skill that touches sensitive operations includes explicit safety rules.
Built on shoulders of giants. Each skill was informed by excellent open-source work. See CREDITS.md for original authors and what I learned from each.
MIT - see LICENSE.
| Tool | What It Does | Deep Dive |
|---|---|---|
| Evolving Lite | Self-improving Claude Code plugin — memory, delegation, self-correction | Blog |
| Kairn | Persistent knowledge graph with context routing for AI | Blog |
| tmux Orchestration | Parallel Claude Code sessions with heartbeat monitoring | Blog |
| UPF | 3-stage planning with adversarial hardening | Blog |
| Quantum Lens | 7 cognitive lenses for multi-perspective analysis | Blog |
| PrimeLine Skills | 5 production-grade workflow skills for Claude Code | Blog |
| Starter System | Lightweight session memory and handoffs | Blog |