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 skills add patleeman/neon-pilot --skill async-attentiongit clone --depth 1 https://github.com/patleeman/neon-pilotWrote 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/patleeman/neon-pilot/async-attention)<a href="https://agentmods.dev/skills/patleeman/neon-pilot/async-attention"><img src="https://agentmods.dev/badge/skills/patleeman/neon-pilot/async-attention/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/patleeman/neon-pilot/async-attention"><img src="https://agentmods.dev/badge/skills/patleeman/neon-pilot/async-attention.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00028 | $0.00815 |
| Opus 5.5 | $0.00011 | $0.00326 |
| Sonnet 5.5 | $0.00006 | $0.00163 |
| Haiku 4.5 | $0.00003 | $0.00081 |
Grade A, and why
async-attention 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 2mo 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 — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Async Attention and Wakeups
Use the smallest scheduling surface that matches the owner of the work. For wait-then-continue requests, use the unified conversation deferred-resume admin command; do not run sleep in bash.
Choose the right surface
| Need | Use | Durable home |
|---|---|---|
| Agent should continue this conversation later | the unified conversation deferred-resume admin command | live queue or deferred resume |
| Unattended automation should run later or recur | scheduled_task |
automation store + run logs |
| Passive async result tied to a thread | surface the owning conversation | conversation/activity |
| Scheduled task result should come back to a thread | scheduled task conversation callback | task log + optional wakeup |
There is no standalone tell-me-later tool. Human “tell me later” requests are same-thread follow-ups unless they need a true app-wide automation.
the unified conversation deferred-resume admin command
Use the unified conversation deferred-resume admin command when this same conversation should continue later.
Actions:
add— queue a follow-uplist— list pending follow-ups for this conversationcancel— cancel a queued follow-up by listedid
For action: "add", always include trigger:
trigger: "after_turn"queues work after the current turn; do not includedelayorat.trigger: "delay"queues a later continuation; include compact duration syntax like30s,10m,2h,4h, or1d.trigger: "at"queues a later continuation for a specific timestamp or human time phrase.
Example:
{
"action": "add",
"trigger": "delay",
"delay": "4h",
"deliverAs": "followUp",
"title": "Check release state",
"prompt": "Wake up and check whether the release is ready. If it is still blocked, requeue instead of asking the user."
}
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.
- 2mo ago First seen · 77 lines · 28 tokens per session scan A a4efe14b3142
async-attention is a skill published in the GitHub repository patleeman/neon-pilot (23 stars, last pushed 3mo ago), licensed MIT. It adds 28 tokens to every session and 815 once invoked, about $0.0001 per session on Opus 5.5. 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.
async-attention compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| async-attentionpatleeman/neon-pilot | Skill | 23 | 3mo ago | MIT | A | 28 | Claude Code, Codex |
| orbit-notionnexu-io/open-design | Skill | 106k | yesterday | Apache-2.0 | A | 117 | Claude Code, Codex |
| agentmailpaperclipai/paperclip | Skill | 111k | yesterday | MIT | A | 30 | Claude Code, Codex |
| Cortexdanielmiessler/LifeOS | Skill | 20k | yesterday | MIT | A | 184 | Claude Code |
| rowboat-spacesrowboatlabs/rowboat | Skill | 18k | yesterday | Apache-2.0 | A | 60 | Claude Code, Codex |
| feishuHmbown/CodeWhale | Skill | 41k | 5d ago | MIT | A | 33 | Claude Code, Codex |
| pinchtab-mcppinchtab/pinchtab | Skill | 11k | 10d ago | MIT | A | 52 | Claude Code, Codex |