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/kirodotdev/kirocrew/babysitnpx skills add kirodotdev/KiroCrew --skill babysitgit clone --depth 1 https://github.com/kirodotdev/KiroCrewWrote 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/kirodotdev/kirocrew/babysit)<a href="https://agentmods.dev/skills/kirodotdev/kirocrew/babysit"><img src="https://agentmods.dev/badge/skills/kirodotdev/kirocrew/babysit.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 | $0.00137 | $0.09970 |
| Opus 5 | $0.00068 | $0.04985 |
| Sonnet 5 | $0.00027 | $0.01994 |
| Haiku 4.5 | $0.00014 | $0.00997 |
Grade A, and why
babysit 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 — 672 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Babysit (same-session monitoring loop)
Overview
monitor_start(message, interval_secs?, max_cycles?) binds a monitoring loop
to your current session. Every interval_secs the message is re-injected
as your next turn. User messages defer a due fire until their turn ends but do
NOT restart the countdown, so the loop stays on schedule even in a session the
user is actively chatting in. You keep the full conversation context, memory,
and tools on every cycle. Loops persist to ~/.kiro/crew/autonudge.json and
survive gateway restarts (the countdown resumes where it left off).
Works from:
| Surface | Binding | Cadence |
|---|---|---|
| Dashboard chat | bare slot key | deadline timer (user turns defer, never reset) |
| Slack thread | slack:<thread_ts> |
fixed interval after each unattended turn |
| Discord DM | discord:{agent}:direct:{user} |
fixed interval after each unattended turn |
autonudge_stop(reason?) stops the loop bound to the current session from
any of those surfaces. monitor_update(message?, interval_secs?, max_cycles?)
revises the loop already bound to this session in place, keeping its cycle
count — use it when the instruction you armed has gone stale, or to raise the
cap on a loop that is still doing useful work.
interval_secs counts between the loop's own cycles
Each delivered cycle's countdown starts when that cycle's turn ends, so
the real cadence is interval_secs + however long each cycle's work takes. A
300s interval with 5-minute checks wakes you roughly every 10 minutes. Size it
for the gap you want between cycles. User messages in the session never
stretch this: a due fire waits for the user's turn to end, then delivers.
You must stop the loop yourself
max_cycles (default 24) is a runaway backstop, not a finish line. A loop
that coasts into its cap did not complete — it ran out of rope, and whatever
it was watching is still unresolved. Evaluate the exit condition every single
cycle and stop deliberately.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 Changed · +7 lines e53e394b7c46
- 4d ago First seen · 665 lines · 137 tokens per session scan A eaf059da8206
babysit is a skill published in the GitHub repository kirodotdev/KiroCrew (3,532 stars, last pushed 2d ago), licensed Apache-2.0. It adds 137 tokens to every session and 9,970 once invoked, about $0.0007 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 2026-08-30.
Other skills, from other repositories
jira-server-pat-cli
通用 Jira Server / Data Center REST CLI 工作流。用于通过 Personal Access Token(PAT)、Cookie 或 Basic Auth 安全地查询和管理 Jira，包括实例探测、项目与字段元数据、JQL、issue CRUD、transition、指派、评论、工时、附件、issue link、watcher、vote、权限和用户查询。适用于自托管 Jira、2FA、SSO、内部 CA、自签证书、context path 和版本差异场景；禁止硬编码实例域名、内部路径、项目 Key、用户名、凭证、自定义字段 ID 或 issue type ID。.
governance-and-sustainability
Keep a project and its maintainers alive over years. Use when deciding how decisions get made, writing GOVERNANCE.md or a code of conduct enforcement process, adding or removing maintainers, handling burnout or stepping away, setting up funding (GitHub Sponsors, Open Collective, a foundation), handling a fork or a…
issue-triage
Build and run a system for handling incoming issues. Use when an issue backlog has become unmanageable, when setting up issue templates or a label taxonomy, when deciding what to close, when configuring stale bots or automation, or when the user asks "how do I deal with all these issues". Covers issue forms, a minimal…
release-engineering
Version, changelog, tag, and publish a release. Use when cutting a release, setting up automated publishing to npm/PyPI/crates.io/Maven/Homebrew, deciding whether a change is major/minor/patch, writing release notes, adopting conventional commits or changesets, planning a 1.0, or handling a bad release that needs to…
contributor-experience
Turn users into contributors and make their first PR succeed. Use when a project gets stars but no contributions, when first-time PRs stall or get abandoned, when writing or improving CONTRIBUTING.md, when setting up a dev environment that works on the first try, or when the user asks "how do I get contributors" or…
open-source-playbook
Entry point and router for open-source work. Use when the user states an open-source goal without naming a specific task — "I want to open source this", "how do I grow this project", "make this repo look professional", "get more contributors", "prepare for v1.0" — or when a request touches several OSS concerns at…