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/isaacsight/kernel/peekaboo-snapshot-actnpx skills add isaacsight/kernel --skill peekaboo-snapshot-actgit clone --depth 1 https://github.com/isaacsight/kernelWrote 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/isaacsight/kernel/peekaboo-snapshot-act)<a href="https://agentmods.dev/skills/isaacsight/kernel/peekaboo-snapshot-act"><img src="https://agentmods.dev/badge/skills/isaacsight/kernel/peekaboo-snapshot-act.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 | $0.00046 | $0.01450 |
| Opus 5 | $0.00023 | $0.00725 |
| Sonnet 5 | $0.00009 | $0.00290 |
| Haiku 4.5 | $0.00005 | $0.00145 |
Grade A, and why
peekaboo-snapshot-act 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 3d 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 — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Peekaboo Snapshot-Act
Every input surface has a right shape. Web is DOM — address elements by ref through Chrome MCP or Playwright. Audio is OSC — address tracks and clips through AbletonOSC's Live Object Model. Native macOS is AX — address controls through the Accessibility tree. Snapshot-and-act replaces screenshot-and-guess: capture the AX tree once, reference labeled element IDs many times, and let the OS resolve coordinates.
When to Use
- Native macOS app automation where the target exposes Accessibility (Notes, Mail, System Settings, Music, Finder, Xcode, Numbers, Keynote, Pages).
- Forms with multiple fields where one snapshot fills them all.
- Repeated interactions inside the same app state where stable element IDs survive across calls.
- Any task where pixel coordinates would drift between window resizes, theme changes, or display scales.
Do not use this skill for:
- Browser tabs — use Chrome MCP (
mcp__claude-in-chrome__*); the DOM is the right surface. - Ableton Live or other audio software with OSC bridges — use
ableton_*tools; OSC is the right surface. - Apps with no Accessibility support — fall through to synthetic input via
mcp__computer-use__*and accept the cost.
Iron Laws
ONE SNAPSHOT, MANY ACTIONS.
ELEMENT ID OVER COORDINATES.
PERFORM-ACTION OVER CLICK.
A snapshot is a contract: while the UI does not change, the IDs are stable. Capture once, act many times, re-snapshot only on visible state change. An element ID survives where coordinates do not — themes shift, windows resize, scroll positions move; elem_169 does not. And when an AX action is named (AXPress, AXShowMenu, AXIncrement), perform it directly; clicking the rendered pixel is a worse approximation of the user's intent.
Five Phases
Phase 1 — Approve & focus
Bring the target app forward and clear the per-app session lock before any Peekaboo call.
app_approve— gate the sensitive-app warning; respect the user's per-app trust state.app_launch— bring the app to the front so AX queries hit the right process.- One-time at the OS level: grant Peekaboo Screen Recording and Accessibility in System Settings → Privacy & Security. Without both,
seereturns an empty tree.
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.
- 3d ago First seen · 94 lines · 46 tokens per session scan A f4333e51bc4b
peekaboo-snapshot-act is a skill published in the GitHub repository isaacsight/kernel (16 stars, last pushed 7d ago), licensed MIT. It adds 46 tokens to every session and 1,450 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
routine-from-chat
Turn a plain-language "do this automatically" request into a routine — either event-triggered ("text me when I get an email from my landlord") or scheduled ("remind me every Friday at 9am"). For third-party events (Gmail, GitHub, Slack…) it confirms with the guardian and then subscribes to the connector trigger so the…
inkbox-cli
Use when running or writing shell commands with the Inkbox CLI (inkbox / @inkbox/cli) for identities, email, mailbox imports, phone, text/SMS, iMessage, A2A task/message history, contacts, notes, contact rules, vault, mailbox storage, mail clients (IMAP/SMTP), phone number, webhook, or signup workflows.
instar-session
Spawn, monitor, and communicate with persistent Claude Code sessions running in the background. Use when a task needs to run without blocking the current session, when the user asks to do something in the background, or when a long-running task needs its own context window. Trigger words: background task, spawn…
status
Mid-session status report. Summarizes what's been achieved so far this session and what work remains, in succinct qualitative prose and bullets with only light technical detail. Read-only — stores nothing, files nothing. Invoke any time to take stock without closing the session.
advogado-criminal-v2
ADVOGADO CRIMINALISTA SENIOR — ESPECIALISTA EM DIREITO PENAL E MARIA DA PENHA workflow skill. Use this skill when the user needs Advogado criminalista especializado em Maria da Penha, violencia domestica, feminicidio, direito penal brasileiro, medidas protetivas, inquerito policial e acao penal and the operator should…
access-request-intake
Collect a complete system-access request from the user and file it via the IT service desk. Use when someone asks to request, grant, or set up system/application access for an employee.