Tax Fraud
Hunts cost everywhere — trims tokens, kills redundant calls, and reports the cheapest path to the same result; invoke via /tax-fraud when you want aggressive (but technically legal) cost optimization of an AI workflow, pipeline, or process.
Install
npx skills add Jakeschincariol/unhinged-claude-skills --skill tax-fraudgit clone --depth 1 https://github.com/Jakeschincariol/unhinged-claude-skills.git /tmp/unhinged-claude-skills
mkdir -p ~/.claude/skills
cp -r /tmp/unhinged-claude-skills/tax-fraud ~/.claude/skills/tax-fraudPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Tax Fraud Every token is a line item, and this mode itemizes ruthlessly. It audits your workflow for waste, deducts everything it can, and reports the cheapest path to the exact same result. Aggressive optimization, fully above board. (Technically.) ## When to use - An AI workflow, pipeline, or process is costing more than it should. - You want the same output for materially less money/tokens/compute. ## Behavior 1. Audit the full cost surface first: every model call, prompt size, retry, and redundant step. 2. Find duplicate or unnecessary calls and eliminate them — cache, memoize, or batch where possible. 3. Trim prompt bloat: cut filler, dedupe context, drop fields the model doesn't need. 4. Right-size the model — flag where a cheaper/smaller model gets the same result. 5. Reduce round-trips: combine calls, stream less, stop paying for output you discard. 6. Quantify the savings concretely — estimate before/after token or dollar cost. 7. Confirm the cheaper path still produces equivalent output before recommending it. ## Output A cost-optimization report: identified waste, the cheaper path to the same result, and concrete before/after estimates — with confirmation that quality is preserved.
Details
- Skill name
tax-fraud- Repository
- Jakeschincariol/unhinged-claude-skills
- Path
tax-fraud/SKILL.md- License
- MIT
- Repo updated
- 27 June 2026
FAQ
What does the Tax Fraud skill do?
Hunts cost everywhere — trims tokens, kills redundant calls, and reports the cheapest path to the same result; invoke via /tax-fraud when you want aggressive (but technically legal) cost optimization of an AI workflow, pipeline, or process.
How do I install the Tax Fraud skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add Jakeschincariol/unhinged-claude-skills --skill tax-fraud. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Tax Fraud popular?
It lives in Jakeschincariol/unhinged-claude-skills, which has 44 GitHub stars and 13 forks.
Related skills
Apollo 13
44Jakeschincariol/unhinged-claude-skills
Works a critical failure under hard constraints — limited options, no clean fix — to get the broken thing home with whatever you have; invoke via /apollo-13 when something important is down, the usual tools are unavailable, and you must improvise a survival fix.
Autopilot Chaos
44Jakeschincariol/unhinged-claude-skills
Takes the whole task and runs it end to end without stopping to ask, leaving the user to review at the finish line; invoke via /autopilot-chaos when you want a fully autonomous run and will inspect the result once it's done.
Basement Hacker
44Jakeschincariol/unhinged-claude-skills
Goes underground to reverse-engineer internals, read the source nobody reads, and explain how something actually works; invoke via /basement-hacker when documentation is missing or lying and you need ground truth from the implementation itself.
Outcome Engineering
38sam-dumont/claude-skills
Use when starting significant features, evaluating build-vs-buy decisions, reframing failed approaches, or when the user mentions "outcome engineering", "o16g", "outcome-first", "what's the outcome", or "frame this as an outcome". Also use proactively when a task description focuses on implementation ("build X", "add Y") without defining the desired outcome or verification criteria.
Triage Validation
44ivRodriguezCA/MobHunt
Load when deciding whether a mobile finding is worth submitting, covering the two-minute triage questions, the Mobile 7-Question Gate, the four post-validation gates, verdict definitions, and the never-submit auto-KILL list.
Rubber Duck
62ElliotJLT/Claude-Skill-Potions
When a user is stuck, frustrated, or describing a problem vaguely, do NOT immediately suggest solutions. First, force structured problem articulation through targeted questions: What did you expect? What happened instead? What have you tried? Only after the problem is clearly defined, propose solutions.
Browse the most popular Claude skills, all finance & crypto skills, or write your own with the skills generator.