From dotnet-artisan
Optimizes .NET development workflow with parallel git worktree sessions, context discipline, plan mode, and verification loops. Helps set up Claude Code for .NET projects and manage token budgets.
How this skill is triggered — by the user, by Claude, or both
Slash command
/dotnet-artisan:dotnet-workflowThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
1. **Parallel over sequential** — Run 3-5 Claude sessions simultaneously with git worktrees. Build a feature in one, fix a bug in another, run tests in a third. Single biggest productivity unlock.
Parallel over sequential — Run 3-5 Claude sessions simultaneously with git worktrees. Build a feature in one, fix a bug in another, run tests in a third. Single biggest productivity unlock.
Plan then execute — Non-trivial tasks: plan mode first, iterate until plan is solid, then auto-accept. A good plan means Claude 1-shots the implementation.
Verification closes the loop — Give Claude a way to prove its work: dotnet build, dotnet test. Without verification, output quality degrades 2-3x.
Automate the repetitive — If done more than once a day: make it a hook, slash command, or subagent. Pre-allow safe permissions. Eliminate friction.
Compound your knowledge — Every correction becomes a rule in MEMORY.md (see Learning System section below). Over time, mistake rate drops as knowledge base grows.
Treat the 200k token window as a budget, not a dumping ground:
# Create parallel work environments
git worktree add ../feature-x feature/x
git worktree add ../bugfix-y bugfix/y
git worktree add ../refactor-z refactor/z
# Each gets its own Claude session
# Clean up when done
git worktree remove ../feature-x
After every code change, run the verification chain:
dotnet build — catches syntax and compilation errorsdotnet test — catches logic and regression errorsPath resolution: The Companion File column lists filenames relative to this skill's
references/directory. Use Glob to locate the file (pattern:**/dotnet-workflow/references/<filename>), then pass the returned absolute path to Read tool.
| Topic | Keywords | Description | Companion File |
|---|---|---|---|
| Plugin verification | test, verify, working, installed, check plugin | How to verify dotnet-artisan is installed and working correctly | references/plugin-verification.md |
Captures and compounds knowledge from user corrections.
1. DETECT — User says "no, use X", "we don't do that", "always/never X here"
2. GENERALIZE — Extract the class-level rule, not the line-specific fix
3. CHECK — Scan MEMORY.md for overlapping rules; update if found
4. STORE — Write categorized rule with rationale
5. CONFIRM — Tell user what was captured: "Added to Memory: ..."
Code Style | Architecture | Naming | Data Access | API Design | Testing | Configuration | Performance
Format: - Rule — rationale
Instincts start as low-confidence hypotheses and follow an observe-hypothesize-confirm cycle:
Store per-project in .claude/instincts.md. Instincts do not transfer between projects.
npx claudepluginhub fenzel999/dotnet-artisan --plugin dotnet-artisanProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.