AIRECEIPTS

Docs ledger

Docs

Start with your first receipt in 60 seconds. Then one guide per command, a reference, and the why-pages.

New here? Get started walks you to your first receipt in under a minute. Everything else answers one question each.

Start here

  • Get started in 60 seconds docs/guide/01-getting-started.md

    aireceipts reads the transcripts your AI coding agent already writes to disk and prints a priced receipt of what a session cost.

  • Install docs/guide/02-install.md

    aireceipts is a Node CLI. It needs Node.js 20 or newer and nothing else — no account, no API key, no daemon.

  • Install the agent hook docs/guide/03-install-hook.md

    Goal: get a short receipt automatically, every time a Claude Code session ends, without running a command yourself.

Using it

  • Read a receipt docs/guide/04-read-a-receipt.md

    Goal: produce and read the receipt for any one session — the newest, or a specific one you pick.

  • Compare two sessions docs/guide/05-compare.md

    Goal: put two receipts next to each other and see the ratio between them — which run cost more, and by how much.

  • Aggregate the week docs/guide/06-week.md

    Goal: total every session from the last seven days, across every agent, with an honest comparison to the week before.

  • Keep the meter running docs/guide/07-statusline.md

    Goal: the meter on *every* Claude Code prompt — model, cost so far, burn rate — without running a command yourself.

  • Set and watch a budget docs/guide/08-budget.md

    Goal: define a spend cap and get a plain signal — a line, and an exit code — when you're near or over it.

  • Fix it next time (handoff) docs/guide/09-handoff.md

    Goal: take the waste a session actually hit and turn it into a paste-ready instruction for your agent — so the next run doesn't repeat it.

  • Use a template docs/guide/10-templates.md

    Goal: render the same receipt in a different visual style — for a screenshot, a share, or just taste.

  • Share and export docs/guide/11-share-and-export.md

    Goal: get a receipt out of the terminal — as an image to share, as data to process, or as a comment on your pull request.

  • Choose an integration docs/guide/15-integrations.md

    Start with the CLI. It is the stable product surface across coding assistants:

Reference

  • Troubleshooting & FAQ docs/guide/12-troubleshooting.md

    Symptom first. Find what you're seeing, then the fix. For question-first answers — how this differs from usage dashboards, what the dollars mean, what leaves your machine — see…

  • FAQ docs/faq.md

    Seeing an error or an odd receipt? That's troubleshooting — symptom-first. This page is question-first. (SPEC-0048.)

  • Export schema (--json / --csv) docs/json-schema.md

    aireceipts can emit a session receipt as machine-readable JSON (--json) or CSV (--csv) instead of the text receipt, for FinOps tooling and spreadsheet ingestion.

  • Statusline — the meter docs/statusline.md

    aireceipts statusline is the meter: one line in your status bar while the agent works — the model that's running, cost so far, burn rate, and the quota countdown — updated as the…

  • PR receipts — attach the building session's receipt to a PR docs/pr-receipts.md

    aireceipts pr attaches the cost receipt of the AI-agent session that built a branch to that branch's pull request, as a single marked comment refreshed on each push.

  • Rolling out receipt checks across a fleet of repos docs/adopt/org-rollout.md

    For one repo, notice-only adoption starts with the workflow caller; automatic receipt posting needs the workflow plus a producer hook for each agent in use.

  • Diagnostics and adoption telemetry docs/telemetry.md

    aireceipts can send small, content-free telemetry events to help us find bugs and understand which CLI features are actually used.

  • Changelog docs/CHANGELOG.md

    All notable changes to aireceipts-cli. Factual, grouped by conventional-commit type (I6: a log, not marketing). Dates are UTC.

Why

  • How pricing is estimated docs/guide/13-pricing.md

    Every dollar on a receipt is an estimate computed locally from token counts and a cited price table — not a feed from your vendor's billing system.

  • How session attribution works docs/guide/14-session-attribution.md

    A receipt is only as trustworthy as its answer to one question: *which session is this?* This page explains what aireceipts treats as a session, how it finds them, and why one can…

  • What a receipt proves — and what it can't docs/trust.md

    aireceipts is local-first: it reads agent transcripts off your disk and prints a receipt.

  • The cost model — how a receipt's numbers are computed, and when it flags docs/cost-model.md

    This is the living account of how aireceipts turns a transcript into priced numbers, per scenario and per agent, and — the part that matters most — exactly when the receipt tells…