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/lukacf/meerkat/hook-authoringnpx skills add lukacf/meerkat --skill hook-authoringgit clone --depth 1 https://github.com/lukacf/meerkatWrote 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/lukacf/meerkat/hook-authoring)<a href="https://agentmods.dev/skills/lukacf/meerkat/hook-authoring"><img src="https://agentmods.dev/badge/skills/lukacf/meerkat/hook-authoring.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 | $0.00020 | $0.00521 |
| Opus 5 | $0.00010 | $0.00260 |
| Sonnet 5 | $0.00004 | $0.00104 |
| Haiku 4.5 | $0.00002 | $0.00052 |
Grade A, and why
Hook Authoring 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 6d 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hook Authoring
Use hooks for runtime observation and policy decisions at typed lifecycle points. Hooks can observe, allow, or deny; they should not become hidden owners of runtime truth.
Hook Points
Meerkat provides 8 hook points in the agent lifecycle:
- RunStarted - When the agent run begins
- PreLlmRequest - Before sending to the LLM
- PostLlmResponse - After receiving the LLM response
- PreToolExecution - Before executing a tool call
- PostToolExecution - After tool execution completes
- TurnBoundary - At the boundary between turns
- RunCompleted - When the agent run completes successfully
- RunFailed - When the agent run fails
Execution Modes
- Foreground hooks run in ascending priority, then registration order. A deny short-circuits later foreground hooks at that point.
- Background hooks run concurrently and must declare
capability: observe. Their decisions are discarded. Use them for logging and analytics, not policy enforcement. - Runtime adapters are in-process handlers, commands, or HTTP endpoints. Each entry has a typed adapter, timeout, point, mode, capability, and priority.
Decision Semantics
Foreground hooks return one of:
- Allow: proceed normally.
- Deny: block the operation with a reason.
- Observe only: return no decision and no patches.
Boundaries
Semantic hook patches are retired. Hooks can observe typed projections and deny through the typed decision shape; provider parameters, assistant text, tool arguments/results, and final run text remain owned by the runtime/tool/LLM authority that produced them.
There is no per-hook failure_policy. Invalid configuration, execution
failure, and timeout are typed engine errors. Foreground failures fail the run;
background failures are recorded as dropped background dispatches and do not
become hook-local denials.
Tool hook projections carry optional ToolProvenance, and LLM-response
projections carry typed provider-native server_tool_content. Classify those
typed fields synchronously instead of parsing display text or tool names.
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.
- 6d ago First seen · 63 lines · 20 tokens per session scan A 7ccde8bf8f78
Hook Authoring is a skill published in the GitHub repository lukacf/meerkat (20 stars, last pushed today), licensed Apache-2.0. It adds 20 tokens to every session and 521 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 2026-08-30.
Other skills, from other repositories
frame-macos-notification
拟真 macOS 通知 banner + app icon + 标题正文, 适合 video overlay / 产品发布预告.
prototype-web
可点击的功能性 Web 原型, 含导航、英雄区、特性区、CTA.
deck-course-module
暖纸背景 + Playfair, 左侧学习目标常驻, 含 MCQ 自测页.
deck-pitch
10 页融资 deck, 白底 + 蓝紫渐变 hero, traction 柱状, $X.XM ask.
waitlist-page
【模板: 等候名单页 / Waitlist】 【意图】为新产品 / 早鸟内测做一张极简等候页。 【布局】.
conversational-ux
Design voice and conversational interfaces — dialog flows, error recovery, and persona. Use when the interface speaks and listens rather than being tapped. For graphical input collection, use form-design.