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 skills/craftos-dev/craftbot/heartbeat-processornpx skills add CraftOS-dev/CraftBot --skill heartbeat-processorgit clone --depth 1 https://github.com/CraftOS-dev/CraftBotWrote 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/skills/craftos-dev/craftbot/heartbeat-processor)<a href="https://agentmods.dev/skills/craftos-dev/craftbot/heartbeat-processor"><img src="https://agentmods.dev/badge/skills/craftos-dev/craftbot/heartbeat-processor.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 | $0.00032 | $0.02684 |
| Opus 5 | $0.00016 | $0.01342 |
| Sonnet 5 | $0.00006 | $0.00537 |
| Haiku 4.5 | $0.00003 | $0.00268 |
Grade A, and why
heartbeat-processor 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 yesterday.
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 — 310 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Heartbeat Processor
Silent background skill for executing scheduled proactive tasks. A single unified heartbeat runs every hour and checks ALL frequencies — hourly, daily, weekly, and monthly tasks are evaluated in one pass.
Trigger Context
You receive a single heartbeat trigger with:
type: "proactive_heartbeat"- The task instruction tells you how many due tasks were found
CRITICAL: Silent Execution (Override Standard Task Rules)
This skill overrides standard task completion rules. Unlike regular tasks:
- NO acknowledgement: Do NOT acknowledge task receipt to user
- NO confirmation: Do NOT wait for user confirmation before ending
- MUST end silently: Use
end_turnimmediately after processing, without user interaction - Can send messages: You can use
send_messagefor tier 1 notifications — notify only, never phrase it as a question that expects an answer - NEVER block on user: Do not end the run on a question at any point
Why? Heartbeat tasks run automatically at regular intervals. Waiting for user confirmation would cause tasks to pile up indefinitely.
CRITICAL: Two Execution Types
When executing proactive tasks, you MUST choose between two execution types:
INLINE Execution
Execute the task directly within this heartbeat session.
Use INLINE when:
- Permission tier is 0 or 1
- Task can be completed with available actions
- No complex multi-step workflows needed
- Task is quick (notifications, searches, summaries)
Examples: Send notification, search information, read and summarize files
SCHEDULED Execution
Schedule the task as a separate session using schedule_task.
Use SCHEDULED when:
- Task requires complex multi-step execution
- Task may need to spawn sub-tasks
- Task requires action sets not available to heartbeat-processor
- Task runs extended operations that shouldn't block heartbeat
Examples: Comprehensive research, multi-file analysis, web scraping, code generation
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.
- yesterday First seen · 310 lines · 32 tokens per session scan A 6e79ec59c112
heartbeat-processor is a skill published in the GitHub repository CraftOS-dev/CraftBot (378 stars, last pushed 2d ago), licensed MIT. It adds 32 tokens to every session and 2,684 once invoked, about $0.0002 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 2026-09-03.
Other skills, from other repositories
artifact-storage-policy
管理 Codex 交付物的專案識別、Google Drive/GitHub 路由、檔名、儲存位置與回讀驗證;凡工作報告、外部保存、交付成果或新 repository 選擇涉及遠端儲存時使用。.
output-contract
任務完成前需要一致交付格式、修改清單、測試結果、待確認事項、或使用者明確要求「完成後列出」時使用。適用於程式碼、設定、文件、skills、同步與多步驟任務收尾;本 Skill 不取代內容 readiness 或風險檢查。.
save-to-notion
將 Codex 對話、回答、決策、工作流程與可重用知識整理到使用者的 Notion 知識 Hub。當使用者明確說「save to notion」,或要求保存、歸檔、整理、分類、版本化、更新到 Notion 時使用;若只是要求產出一般報告或文件,除非另有指定,不要觸發 Notion 實際寫入。.
delivery-manifest
每次產生、修改、上傳或寄送檔案後,產生統一的交付物索引 (Delivery Manifest),並嚴格控管 Google Drive 上傳路徑。.
sales-daily-briefing
Start your day with a prioritized sales briefing. Works standalone when you tell me your meetings and priorities, supercharged when you connect your calendar, CRM, and email. Trigger with "morning briefing", "daily brief", "what's on my plate today", "prep my day", or "start my day".
productivity-start
Initialize the productivity system and open the dashboard.