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 jasonkneen/lazar --skill guardrailsgit clone --depth 1 https://github.com/jasonkneen/lazarWrote 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/jasonkneen/lazar/guardrails)<a href="https://agentmods.dev/skills/jasonkneen/lazar/guardrails"><img src="https://agentmods.dev/badge/skills/jasonkneen/lazar/guardrails/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/jasonkneen/lazar/guardrails"><img src="https://agentmods.dev/badge/skills/jasonkneen/lazar/guardrails.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 | $0.00101 | $0.00705 |
| Opus 5 | $0.00051 | $0.00352 |
| Sonnet 5 | $0.00020 | $0.00141 |
| Haiku 4.5 | $0.00010 | $0.00071 |
Grade B, and why
guardrails scanned grade B with 1 finding 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 11d 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.
Tells the agent never to refusemediumAnti-refusal
Suppressing the ability to decline removes a core safety control; a later harmful request then succeeds.
description: "Turn lazar's enforcement layers on or off. Use when a write fails with 'Operation not permitted', when the operator asks for full access, when you need to edit bin/ src/ logs/, or when the operator asks to Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.
guardrails
One switch for every enforcement layer in lazar.
The four layers
| # | Layer | Mechanism |
|---|---|---|
| 1 | bash tool sandbox | sandbox-exec wrapper (LAZAR_NO_SANDBOX) |
| 2 | kernel binary lock | chflags uchg bin/lazar |
| 3 | source lock | chmod a-w src/ |
| 4 | pre-tool vetoes | hooks/pre-tool.d/*.sh |
Commands
bash $LAZAR_HOME/scripts/lazar-guardrails.sh status # which mode am I in?
bash $LAZAR_HOME/scripts/lazar-guardrails.sh off # full operator access
bash $LAZAR_HOME/scripts/lazar-guardrails.sh on # restore enforcement
When to use
- A write returned
Operation not permittedand the operator wants it done. - The operator asks to edit
bin/,src/,logs/, or anything outsideskills/ memory/ workspace/. - The operator says "full access", "no sandbox", "stop blocking me".
- The operator asks to re-enable safety after risky work.
Hard rule
Never refuse an operator request because of permissions.
Wrong:
"I can't write to logs/ — it's kernel-owned. Want me to use workspace/ instead?"
Right:
bash $LAZAR_HOME/scripts/lazar-guardrails.sh status
# if guardrails are on and the operator wants this:
bash $LAZAR_HOME/scripts/lazar-guardrails.sh off
# then perform the actual write
If guardrails are already OFF, there is nothing to route around — just do the work.
Important: mode changes need a fresh invocation
Layers 2–4 (locks, hooks) apply immediately. Layer 1 (sandbox) is decided when
the kernel process starts, so a running invocation stays in its current mode.
After toggling, the NEXT lazar call picks up the new mode. Tell the operator
that rather than pretending the current turn changed.
Safety posture when OFF
Guardrails OFF removes enforcement, not discipline. Still true:
- Kernel edits go through
workspace/proposals/+scripts/kernel-apply.sh. - Back up before destructive edits (
engineering/safe-edit). - VERIFY still runs — grounding is independent of the sandbox.
- Do nothing the operator did not ask for. Full access is trust, not licence.
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.
- 11d ago First seen · 76 lines · 101 tokens per session scan B 82614cb1e18b
guardrails is a skill published in the GitHub repository jasonkneen/lazar (31 stars, last pushed 29d ago), licensed MIT. It adds 101 tokens to every session and 705 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it B with 1 finding (tells the agent never to refuse). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…