metago-decision-lock

metago-decision-lock is a skill for Claude Code, Codex from metago-ai/metagolifeform. It costs 81 tokens per session (1,189 once invoked), scanned A, original, MIT.

A decision-checking procedure for an AI agent. It checks the user's intent, tracks whether actions stay aligned with it, checks the meaning of the response, and looks for placeholders or made-up information.

In plain words
What is it for?
Use it before responses, code execution, file changes, or system configuration changes when the agent must verify that its work is complete and aligned.
Why use it?
It reduces the risk of unclear instructions, off-target actions, unfinished output, invented APIs, or unsupported data.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: mentions Codex.

Good fit Use it before responses, code execution, file changes, or system configuration changes when the agent must verify that its work is complete and aligned.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/metago-ai/metagolifeform/metago-decision-lock
Install

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.

Any agent
npx skills add metago-ai/metagolifeform --skill metago-decision-lock
Clone the repo
git clone --depth 1 https://github.com/metago-ai/metagolifeform

Made for: Claude Code, Codex.

Wrote 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.

agentmods badge for metago-decision-lock

README.md
[![agentmods](https://agentmods.dev/badge/skills/metago-ai/metagolifeform/metago-decision-lock/github.svg)](https://agentmods.dev/skills/metago-ai/metagolifeform/metago-decision-lock)
Your own site
<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.

agentmods 80×15 button for metago-decision-lock

Your own site · 80×15
<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>
Per session 81 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,189 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00081 $0.01189
Opus 5 $0.00041 $0.00594
Sonnet 5 $0.00016 $0.00238
Haiku 4.5 $0.00008 $0.00119

Measured 12d ago against content hash c038424235f7, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

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 12d 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.

plugins/agent-plugins-1.0.0/skills/metago-decision-lock/SKILL.md · 133 lines

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

  • metago-meta-evolve 协同:决策锁未通过时触发元进化
  • metago-self-check 协同:决策锁是self-check的前置关卡
  • metago-critique 协同:决策锁通过后触发批判性分析
  • metago-fact-check 协同:关卡4与事实核查联动
Changes

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.

  1. 12d ago First seen · 133 lines · 81 tokens per session scan A c038424235f7

Subscribe to this mod's changes

metago-decision-lock is a skill published in the GitHub repository metago-ai/metagolifeform (4 stars, last pushed 10d ago), licensed MIT. It adds 81 tokens to every session and 1,189 once invoked, about $0.0004 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 2026-08-31.