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 SuperChason/ontology-driven-ai-data-management-skills --skill risk-based-agent-action-modesgit clone --depth 1 https://github.com/SuperChason/ontology-driven-ai-data-management-skillsWrote 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/superchason/ontology-driven-ai-data-management-skills/risk-based-agent-action-modes)<a href="https://agentmods.dev/skills/superchason/ontology-driven-ai-data-management-skills/risk-based-agent-action-modes"><img src="https://agentmods.dev/badge/skills/superchason/ontology-driven-ai-data-management-skills/risk-based-agent-action-modes/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/superchason/ontology-driven-ai-data-management-skills/risk-based-agent-action-modes"><img src="https://agentmods.dev/badge/skills/superchason/ontology-driven-ai-data-management-skills/risk-based-agent-action-modes.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 | $0.00075 | $0.01255 |
| Opus 5 | $0.00037 | $0.00628 |
| Sonnet 5 | $0.00015 | $0.00251 |
| Haiku 4.5 | $0.00007 | $0.00126 |
Grade A, and why
risk-based-agent-action-modes 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 3d 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
基于风险的三类行动模式选择
方法骨架
- 根据风险、规则确定性、可恢复性和协作复杂度选择Agent行动模式。
- 低风险、规则清晰、结果可回滚的任务可采用指令直驱。
- 高风险、主观判断或强合规任务设置分层人工介入。
- 跨系统、多角色、多阶段任务采用业务协同,由Agent与人共同编排。
- 授权至少区分自动、需确认和禁止,并明确责任与审计。
- 人工调整和执行反馈需要回写,持续校正模式选择规则。
触发场景
用户会在什么情境下需要这个 Skill
- 要决定Agent能否自动执行某个业务动作
- 需要设计人工审批或接管点
- 单Agent与多Agent协同方案之间需要选择
语言信号
- “这个动作能自动执行吗”
- “哪里需要人工介入”
- “该用单Agent还是多Agent协同”
- 英文信号:human in the loop, action autonomy, multi-agent mode
与相邻 Skill 的区分
- 与
action-contract-execution-feedback-loop:本 skill 选择自动、人工或协同模式;Action契约 skill 定义具体动作怎样执行和反馈。 - 与
ontology-ai-application-pattern-selection:应用模式选择决定业务形态;本 skill 进一步决定其中每类行动的控制方式。
执行步骤
按当前任务选择必要步骤;已有可靠成果直接复用:
-
登记行动
- 动作:列出动作对象、影响范围、权限、频次、可恢复性和失败后果。
- 完成标准:每个动作有业务责任人和目标系统。
-
评估四维
- 动作:评估风险、规则确定性、可恢复性、协作复杂度,并记录证据。
- 完成标准:四维等级和不可接受后果明确。
-
选择模式
- 动作:低风险清晰任务用直驱;高风险或主观任务用人工介入;跨系统多角色任务用业务协同。
- 完成标准:每个动作只有一个主模式,升级与降级条件明确。
-
配置授权
- 动作:设置自动、确认、禁止,以及审批人、超时、接管和审计要求。
- 完成标准:高风险写操作没有默认自动权限。
-
回写反馈
- 动作:根据成功、人工修改、失败和异常更新规则与模式。
- 完成标准:形成行动模式矩阵和反馈更新机制。
固定输出
- 行动清单:行动编号、对象、业务影响、目标系统、责任人、频次、权限和可恢复性
- 行动风险评估矩阵:风险、规则确定性、可恢复性、协作复杂度、失败后果、证据和不可接受结果
- 行动模式矩阵:行动、自动执行、人工确认、业务协同或禁止的主模式及理由
- 授权与人工介入表:授权角色、阈值、确认点、审批人、超时、接管、终止和审计要求
- 模式升级、降级与停止条件表
- 执行结果、人工修正、异常和模式调整反馈表
- 剩余风险、不可自动化项与下游 Action 契约输入清单
每个行动只保留一个主控制模式,高风险写操作没有默认自动权限,模式变更保留原因和审计记录。
使用边界
不要在以下情况使用
- 只需描述具体API参数和错误码
- 业务风险和责任人尚未确定
- 简单任务为了展示技术而强行拆成多Agent
常见失败模式
- 高风险行动采用规则直驱:决策正确性与执行授权被合并,缺少人工裁决和后果控制,单点错误直接转为现实损失。
- 简单任务过度采用多Agent协同:任务分解收益不足以抵消通信、路由、共享状态和冲突处理成本。
使用折扣与复核要求
- 风险等级属于企业治理决定,三类行动模式不能代替本组织授权制度。
- 大模型生成形式结构无法直接证明业务语义正确,生产使用需保留专家确认、工具校验、真实用例和审计记录。
相关 Skills
depends-on→fact-reason-goal-explainable-decision。
审计信息
- 历史验证:v0.1.0 路由测试 6/6;v0.4.0 已通过输出契约结构校验,跨平台行为继续按版本抽样
- 首次公开版本:2026-08-21
- 来源说明:方法框架受《本体驱动的 AI 数据管理》启发;仓库不包含原书正文。
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.
- 3d ago Changed 7586e8e5bdb0
- 7d ago Changed · +7 lines 406f5fa963aa
- 11d ago First seen · 97 lines · 75 tokens per session scan A 7fc032a5a952
risk-based-agent-action-modes is a skill published in the GitHub repository SuperChason/ontology-driven-ai-data-management-skills (10 stars, last pushed 4d ago), licensed MIT. It adds 75 tokens to every session and 1,255 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.
Other skills, from other repositories
graph-mutation-plan
Cookbook for composing an applygraphmutations plan — stable entitykey patterns, the canonical label/edge vocabulary, evidence/invalidation/confidence discipline, and a worked example. Load this when building a non-trivial mutation plan.
open-ontologies
AI-native ontology engineering using 50+ MCP tools backed by an in-memory Oxigraph triple store. Build, validate, query, and govern RDF/OWL ontologies with a generate-validate-iterate loop. Use when building ontologies, knowledge graphs, RDF data, SPARQL queries, BORO/4D modeling, SHACL validation, clinical…
report-generation
A workflow for generating data-analysis reports as interactive HTML with charts. It is intended for trend, statistics, monthly, weekly, and other reports, using database queries and ECharts, a web charting library.
business-overview
A business performance analysis workflow that uses sales, finance, inventory, and customer data to describe how a company is operating.
enterprise-sales
A Chinese-language workflow for preparing sales visits and solution documents for government and enterprise customers. It covers checking customer records, understanding the industry, finding suitable products, and generating a Word proposal.
mykg
Run mykg knowledge-graph commands inside Claude Code from one slash command /mykg. The user describes intent in natural language (extract, append, sync, resume, approve, walkthrough, parse-docs, fetch-web, query); the skill parses intent, builds the right mykg CLI command from the live --help output, confirms, runs…