Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/emaraschio/cursor-commands/agent-work-receiptgit clone --depth 1 https://github.com/emaraschio/cursor-commandsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/emaraschio/cursor-commands/agent-work-receipt)<a href="https://agentmods.dev/commands/emaraschio/cursor-commands/agent-work-receipt"><img src="https://agentmods.dev/badge/commands/emaraschio/cursor-commands/agent-work-receipt.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00027 | $0.00644 |
| Opus 5 | $0.00014 | $0.00322 |
| Sonnet 5 | $0.00005 | $0.00129 |
| Haiku 4.5 | $0.00003 | $0.00064 |
Grade A, and why
agent-work-receipt scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Overview
Produce a conservative agent work receipt after completed work: finished output, time estimates, review burden, risks, and value judgment. Retrospective only; no new implementation in the same turn. Full workflow: .cursor/skill-contracts/agent-work-receipt/SKILL.md (user install: ~/.cursor/skill-contracts/agent-work-receipt/SKILL.md).
Defaults
| Setting | Default |
|---|---|
| Scope | Full session unless user narrows in the message |
| Time fields | Agent estimates, labeled as estimates |
| Evidence | Conversation + read-only git (status, diff, log) when repo present |
| Execution | Retrospective only, no new implementation in same turn |
| Examples | Generic names only (service-a, repo1) |
Steps
- Read
.cursor/skill-contracts/agent-work-receipt/SKILL.mdfor the full agent contract; if that file is missing, read~/.cursor/skill-contracts/agent-work-receipt/SKILL.md. - Execute phases in order (Intake → Evidence → Inventory → Estimates → Review & risk → Value → Deliver).
- Report the six-section receipt in chat using the skill template.
Anti-patterns
- Count only landed work, not drafts. Trigger: tallying finished output. Wrong: counting drafts, ideas, or abandoned attempts as completed work. Correct: list only artifacts that actually landed and stay conservative. Reason: a receipt that counts unused output overstates value and misleads the decision it informs.
- Stay retrospective; start no new work. Trigger: writing the receipt for a finished session. Wrong: starting new implementation, commits, or destructive commands in the same turn. Correct: account for completed work only and defer new work to a separate turn. Reason: a receipt that mutates the repo it is measuring corrupts its own evidence.
- Label time as estimates, never measured fact. Trigger: reporting human baseline and agent-assisted time. Wrong: presenting invented durations as measured elapsed time, or inflating savings to match a request. Correct: give estimates labeled as estimates and backed by evidence. Reason: fabricated time figures turn the receipt into marketing rather than accounting.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 46 lines · 27 tokens per session scan A cb2629a98485
agent-work-receipt is a command published in the GitHub repository emaraschio/cursor-commands (9 stars, last pushed 29d ago), licensed MIT. It adds 27 tokens to every session and 644 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
adhoc-flow
Workflow for the rest of tasks: lightweight documentation, build, track, synchronize, etc.
prep
Prepare the meeting. One page from the record.
start-0-7.en
Auto-run by AI: Run uv run python tools/setupprogress.py show to display the current setup progress.
setup-m365cli
Lesson command — M365 CLI セットアップ.
start-11-3.en
Lesson 11-3: News Fetching → Email/Slack Distribution!
start-11-3
Command "start-11-3" from minicoohei/ai-agent-camp, covering 🎓 lesson 11-3: ニュース取得→メール/slack配信ワークフロー, 📍 このセッションでやること, 🎯 準備チェック, 🚀 step 1: ニュース取得スクリプトの作成 and rss フィード url（例: はてなテクノロジー）.