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 HUAFIRE777/autorunne-grill --skill autorunne_grillgit clone --depth 1 https://github.com/HUAFIRE777/autorunne-grillWrote 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/huafire777/autorunne-grill/autorunne_grill)<a href="https://agentmods.dev/skills/huafire777/autorunne-grill/autorunne_grill"><img src="https://agentmods.dev/badge/skills/huafire777/autorunne-grill/autorunne_grill/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/huafire777/autorunne-grill/autorunne_grill"><img src="https://agentmods.dev/badge/skills/huafire777/autorunne-grill/autorunne_grill.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 | $0.00054 | $0.02944 |
| Opus 5.5 | $0.00022 | $0.01178 |
| Sonnet 5.5 | $0.00011 | $0.00589 |
| Haiku 4.5 | $0.00005 | $0.00294 |
Grade A, and why
autorunne-grill 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.
This is a copy
100% identical to autorunne-grill — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 307 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Autorunne Grill
Overview
Autorunne Grill is a project-aware version of the classic grill-me skill.
The original grill-me asks relentless questions about a plan. Autorunne Grill does something more practical for real development: it first reads the repo's Autorunne project memory, then asks only the questions that are still necessary to keep the next feature change inside a safe boundary.
Use it as a pre-change safety gate for beginner-friendly AI development. The user should be able to say "add a login feature" or "change the checkout page" without knowing how to manage engineering workflow details. The agent reads Autorunne, checks the current project state, narrows the change, records the task, implements the smallest safe slice, verifies it, and finishes cleanly.
When to Use
Use this skill when all three are true:
- the repo has
.autorunne/,.autorunne/views/, or Autorunne-generated handoff files; - the user asks to add, change, remove, redesign, or "just tweak" a feature;
- the change could affect product behavior, project direction, data, UI flow, tests, or integration points.
Good triggers:
- "帮我加一个登录功能"
- "改一下支付页面"
- "把线索导出成 CSV"
- "优化自动记录逻辑"
- "重做这个页面"
- "加一个会员中心"
- "grill this plan, but use Autorunne state first"
Do not use this for:
- pure explanation with no project change;
- simple typo fixes with no behavior impact;
- one-off shell questions unrelated to the repo;
- tasks where the user explicitly asks for no planning and the change is obviously tiny.
Core Rule
Do not grill the user from a blank slate.
First read Autorunne state and the relevant code. Then ask the fewest questions needed to safely proceed.
If a question can be answered by reading .autorunne/, generated views, tests, README, package files, routes, schema files, or existing code, answer it yourself instead of asking the user.
Installation Modes
Prefer a repo-local install when the goal is to make one project carry the rule with it. In 0.1.4+, the short command auto-installs into the current Autorunne-backed repo:
What ships with it
3 files 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.
- 2mo ago First seen · 307 lines · 54 tokens per session scan A 5d875b9320e3
autorunne-grill is a skill published in the GitHub repository HUAFIRE777/autorunne-grill (2 stars, last pushed 5mo ago), licensed MIT. It adds 54 tokens to every session and 2,944 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. It is 100% identical to autorunne-grill, differing in 0 lines, and is treated as a copy.
Other skills, from other repositories
lockin
Lock in on a goal that can be checked and work autonomously until it's provably met or a cap is hit, looping act → prove → judge. Use when the user runs /lockin, says "lock in", "keep going until…", "don't stop until…", or hands over a task with a clear finish line (tests pass, build green, queue empty).
pre-flight
Pre-production code review against a user-chosen healthy point. Use when asked to review a branch before merge, whether it's safe to ship, or what blocks a PR.
preflight
Pre-production code review against a user-chosen healthy point. Use when asked to review a branch before merge, whether it's safe to ship, or what blocks a PR.
auto-implement
Autonomously implement a masterplan plan phase-by-phase. A central orchestrator spawns one herdr agent per phase (via the implement-plan skill), watches it to completion, commits, and moves to the next phase until the plan is done. Use after grilling + create-plan when the user wants a plan fully implemented…
create-plan
Create a multi-phase implementation plan using tracer-bullet vertical slices, and submit via the createplan masterplan tool. Use when user wants to create an implementation plan or mentions planning.
handoff
Compact the current conversation into a handoff document for another agent to pick up.