receipts

Generate a personal Claude Code usage and impact report from local session transcripts.

1|Updated Jul 6, 2015
One-click install
npx skills add https://github.com/ksolomon/dotfiles --skill receipts-ksolomon
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: receipts
Source: https://github.com/ksolomon/dotfiles/tree/main/AI/.claude/plugins/marketplaces/claude-plugins-official/plugins/receipts/skills/receipts
Command: npx skills add https://github.com/ksolomon/dotfiles --skill receipts-ksolomon

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Developers who use Claude Code often struggle to explain or justify their usage and spend to managers, or to answer "what have I actually been using this for?" This Skill mines local session transcripts and git history to produce a concrete, evidence-based impact report without manual data gathering. ## Core Features & Use Cases - Local transcript mining: A Node script scans ~/.claude/projects session logs offline, aggregating sessions, prompts, active days, files touched, and per-project spend share with zero extra API calls. - Git cross-referencing: Correlates Claude Code activity with local git history to count commits that actually carry Claude-assisted work, avoiding inflated or misleading numbers. - Dual output: Produces a markdown report for pasting into docs plus a self-contained HTML receipt with an Export CSV button, saved to your home directory. - Use Case: Before a performance review, run the report for the last quarter to show your manager which projects consumed Claude Code time, what shipped, and how many PRs were opened. ## Quick Start Ask Claude to run /receipts for the past month to generate your Claude Code impact report.

Frequently Asked Questions about receipts

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a Claude Code usage report?▼

Run the /receipts command with a time period like week, month, quarter, or year. The skill mines your local session transcripts at ~/.claude/projects, cross-references git history, and writes a markdown report plus an HTML receipt to your home directory.

Does the receipts report send my data anywhere?▼

No network calls are made during mining; the script only reads local JSONL transcripts and runs local git commands. The only data sent to the model is a small JSON summary with aggregate counts and project names, never code or conversation content.

Can I scope the report to a single project?▼

Yes, pass a project name or substring as an argument, such as /receipts for myproject. The miner filters on the resolved project name case-insensitively and scopes the entire report, including totals, to matching projects.

Why does the report show no dollar cost figures?▼

Dollar costs computed from local token counts would be inferred rather than measured and would not match your actual bill. The report instead shows each project's percentage share of relative compute, which is a stable, defensible metric.

What does a null commit count mean in the report?▼

A null commit count means either git errored while reading the repository or no commits carrying Claude Code's work landed in that project. The report distinguishes these cases using the gitUnavailable flag rather than reporting a flat zero.