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 metago-ai/metagolifeform --skill metago-decision-lockgit clone --depth 1 https://github.com/metago-ai/metagolifeformWrote 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/metago-ai/metagolifeform/metago-decision-lock)<a href="https://agentmods.dev/skills/metago-ai/metagolifeform/metago-decision-lock"><img src="https://agentmods.dev/badge/skills/metago-ai/metagolifeform/metago-decision-lock/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/metago-ai/metagolifeform/metago-decision-lock"><img src="https://agentmods.dev/badge/skills/metago-ai/metagolifeform/metago-decision-lock.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 | $0.00081 | $0.01189 |
| Opus 5.5 | $0.00032 | $0.00476 |
| Sonnet 5.5 | $0.00016 | $0.00238 |
| Haiku 4.5 | $0.00008 | $0.00119 |
Grade A, and why
metago-decision-lock 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 1mo 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.
What it actually says
决策锁(Decision Lock)
此技能实现"决策锁"——在AI Agent的思考-行动链中设置强制校验节点,确保每个决策都经过严格验证,杜绝幻觉、越界、不一致输出。
触发条件
- 每次输出前强制触发
- 涉及代码执行、文件修改、系统配置时强制触发
- 开发者任务强制触发
- 检测到引用占位符时强制触发
四道校验关卡
关卡1:意图验证层(IVL - Intent Verification Layer)
验证用户意图是否合法、清晰、可执行:
检查项:
1. 意图是否明确(非模糊指令)
2. 意图是否合法(不违反法律、伦理、安全)
3. 意图是否在能力边界内
4. 意图是否有足够上下文
判定:
- 全部通过 → 进入关卡2
- 意图模糊 → 请求澄清
- 意图非法 → 拒绝并说明原因
- 超出能力 → 触发元进化
关卡2:意图谱系追踪(ILT - Intent Lineage Tracing)
追踪意图的完整谱系,确保行动链一致:
检查项:
1. 当前行动是否与原始意图一致
2. 行动链是否有断裂
3. 是否存在意图漂移(目标偏移)
4. 谱系是否可完整溯源
判定:
- 谱系一致 → 进入关卡3
- 谱系断裂 → 重新对齐意图
- 意图漂移 → 纠正回原始意图
关卡3:语义输出门(OSG - Output Semantic Gate)
验证输出语义是否匹配意图:
检查项:
1. 输出内容是否回应了意图
2. 输出语义是否与意图语义匹配
3. 是否存在语义偏移
4. 输出是否完整(无缺失环节)
判定:
- 语义匹配 → 进入关卡4
- 语义偏移 → 重新生成
- 输出缺失 → 补全闭环
关卡4:内容完整性校验
检测引用占位符与幻觉输出:
检查项:
1. 引用占位符检测({{}}、[]、TODO、FIXME、...)
2. 幻觉检测(虚构的API、不存在的函数、伪造的数据)
3. 引用完整性(所有引用是否可溯源)
4. 数据真实性(所有数据是否有来源)
判定:
- 无占位符、无幻觉 → ✅ 通过,允许输出
- 检测到占位符 → ❌ 拒绝,重新生成
- 检测到幻觉 → ❌ 拒绝,重新生成
负样本库
参考案例 NEG_20260508_002:
- 问题:输出中包含未替换的引用占位符
- 原因:未通过关卡4内容完整性校验
- 修正:强制重新生成,直至占位符全部替换
输出格式
决策锁校验完成后,附加校验报告:
【决策锁校验】
关卡1 意图验证(IVL):✅ 通过
关卡2 意图谱系追踪(ILT):✅ 通过
关卡3 语义输出门(OSG):✅ 通过
关卡4 内容完整性:✅ 通过
校验结果:✅ 允许输出
若任一关卡未通过:
【决策锁校验】
关卡1 意图验证(IVL):✅ 通过
关卡2 意图谱系追踪(ILT):❌ 未通过(意图漂移)
校验结果:❌ 拒绝输出,重新对齐意图
与其他技能的协同
- 与
metago-meta-evolve协同:决策锁未通过时触发元进化 - 与
metago-self-check协同:决策锁是self-check的前置关卡 - 与
metago-critique协同:决策锁通过后触发批判性分析 - 与
metago-fact-check协同:关卡4与事实核查联动
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.
- 1mo ago First seen · 133 lines · 81 tokens per session scan A c038424235f7
metago-decision-lock is a skill published in the GitHub repository metago-ai/metagolifeform (5 stars, last pushed today), licensed MIT. It adds 81 tokens to every session and 1,189 once invoked, about $0.0003 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 2026-08-31.
Other skills, from other repositories
todo-for-ai-agent-runtime
Runtime integration skill for external agents connecting to Todo for AI over HTTP API Token / Agent Key, including auth, task lease, events, commit, and task logs.
monetization-guide
A Chinese-language guide for creating business and pricing plans. It first classifies the situation, such as launching a new paid product, changing prices, or exploring a new way to earn revenue.
wechat-article-scraper
A batch tool for downloading WeChat public-account articles from their URLs as Markdown files. It also downloads the articles' images and links them locally.
karpathy-guidelines
Tool-agnostic behavioral guidelines for AI coding assistants. Use when writing, reviewing, debugging, or refactoring code to reduce overengineering, surface ambiguity, make surgical changes, and define verifiable success criteria.
prototype-web
A clickable, high-fidelity web product prototype with navigation, a hero section, feature cards, steps, social proof, and optional pricing. It is designed to resemble a finished landing page while remaining a prototype.
loop-engineering
Shared loop-engineering reference for COG skills - the agent loop, deterministic verifiers, termination conditions, in-loop context management, and named patterns. Invoke when designing or debugging a skill that iterates (search-verify-retry, scan-until-dry, fetch-retry-gate).