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 DongLiStudio/personal-agent-foundation --skill decide-next-actiongit clone --depth 1 https://github.com/DongLiStudio/personal-agent-foundationWrote 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/donglistudio/personal-agent-foundation/decide-next-action)<a href="https://agentmods.dev/skills/donglistudio/personal-agent-foundation/decide-next-action"><img src="https://agentmods.dev/badge/skills/donglistudio/personal-agent-foundation/decide-next-action/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/donglistudio/personal-agent-foundation/decide-next-action"><img src="https://agentmods.dev/badge/skills/donglistudio/personal-agent-foundation/decide-next-action.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 | $0.00160 | $0.01815 |
| Opus 5 | $0.00080 | $0.00907 |
| Sonnet 5 | $0.00032 | $0.00363 |
| Haiku 4.5 | $0.00016 | $0.00181 |
Grade A, and why
decide-next-action 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 10d 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.
How it starts
The opening of the file, as written. The whole thing — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.
决定下一步行动
从用户现有系统中取得足够而不过量的实时证据,给出一个明确、可立即执行的下一步。不要把结果退化为待办清单、完整日程或泛泛建议。
职责边界
- 默认由{{GENERAL_ASSISTANT_PROJECT}}的“个人时间管理员”承接;已有等价岗位或窗口时复用,不新增重复岗位。暂时无法调用长期岗位时,由当前会话完成,不得因此阻塞。
- 本 Skill 只做即时决策和只读建议,不修改任何来源。
- 用户要安排今天、未来几天或本周时,调用
personal-schedule-planner。 - 用户要创建、更新或记录飞书任务时,调用
feishu-task。 - 用户只问某个明确项目内部的下一步时,以该项目为主,其他来源只用于发现硬冲突;不要无意义地全局采集。
工作流程
1. 确定决策窗口
读取当前带时区时间,识别下一项固定日程和此前对话中用户刚刚说明的地点、精力、可用时长或临时要求。
- 默认回答“从现在开始先做什么”,不自动生成全天时间表。
- 默认观察未来七天的截止、依赖和冲突,但只选择当前一个行动。
- 只有地点、精力或可用时长会实质改变首选时,才询问一个最小问题;否则明确合理推定并继续。
2. 只读采集
需要全局判断且环境支持会话内临时子智能体时,并行启动三条只读分支;规模很小或无并发能力时顺序执行,保持相同范围。
飞书
- 读取 GLOBAL 的
LARK_PROFILES.md,覆盖其中全部治理层 Profile,不以当前 active Profile 代替全量范围。 - 每个 Profile 显式使用对应身份,读取当前用户负责或实际需要执行的未完成任务、开始/截止时间、可靠工作量,以及当前至观察窗口内的日历占用。
- 不把仅关注、由他人执行或只是等待反馈的任务误判成用户当前行动;不默认读取无关附件或任务全文。
Obsidian
- 先按 GLOBAL 的
OBSIDIAN_LINK.md定位并读取仪表盘.canvas。 - 只对与候选行动有关的 Wikilink、Canvas 文件节点和受限 Dataview 查询范围选择性读取;需要
.base时按obsidian-bases规则理解。 - 不递归遍历 Vault,不写入 Obsidian,不把愿景或长期方向直接当作当前承诺。
项目
- 从 GLOBAL 的
PROJECTS.md获取活跃项目。 - 读取候选项目的
AGENTS.md、README.md、STATUS.md;只有需要验证明确下一步、阻塞或验收标准时,才读取具体任务文件。 - 以项目实时文件为事实,不用历史摘要代替当前状态,不对项目目录做递归普查。
三条分支使用同一个采集截止时间,并回报实际覆盖、候选事项、固定冲突、风险和准确错误。主 Agent 等待所需分支回报后统一去重和判断;来源缺失会改变结论时,先安全补采,仍失败则说明缺口和置信度。
3. 归一化候选事项
- 以飞书任务 GUID、项目路径、日历事件 ID 和明确交付语义去重;同一事项跨来源出现时保留全部证据,但只形成一个候选项。
- 区分“用户现在可执行”“等待他人”“固定承诺”“长期方向”“仅供参考”。只有第一类进入行动选择,固定承诺用于限制时间窗口。
- 每个候选至少保留:来源、项目、截止、依赖、阻塞对象、建议投入时长及置信度、战略价值、当前环境适配度和可验证完成标准。
- 没有可靠时长时给出保守估时并标明低置信度,不把“尽快”“有空”自行转换为硬截止。
4. 做出选择
完整读取 下一步决策模型。先排除当前不可执行、非用户职责、已被固定日程挤占或证据不足以行动的事项,再结合:
- 已存在且即将发生的固定承诺。
- 已逾期或临近截止的用户责任。
- 会阻塞他人、解除关键依赖或降低重大风险的事项。
- 用户刚刚明确强调的目标与高战略价值推进。
- 当前可用窗口、精力、地点、所需工具和切换成本。
- 能否在建议时间内产生清晰、可验收的阶段结果。
不要使用机械总分替代判断。最终必须选择一个首选;只有证据确实无法区分且一个最小问题会改变结论时才暂停询问。
5. 给出结论
回答应短而具体,先给结论:
现在建议先做:<一个行动>
建议投入:<时长或直到下一固定日程>
完成标准:<本轮可验收结果>
为什么是它:<最关键的 2–4 条依据>
开始动作:<第一步,能够立刻执行>
置信度:高 / 中 / 低;<必要推定>
备选:
1. <只在首选暂时无法执行时使用>
2. <可选;没有合格备选时不硬凑>
What ships with it
2 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.
- 10d ago First seen · 106 lines · 160 tokens per session scan A 0404be94b458
decide-next-action is a skill published in the GitHub repository DongLiStudio/personal-agent-foundation (11 stars, last pushed 23d ago), licensed Apache-2.0. It adds 160 tokens to every session and 1,815 once invoked, about $0.0008 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 2026-08-30.
Other skills, from other repositories
a0-development
Development guide for extending Agent Zero from current source and DOX. Use for framework architecture, tools, extensions, API/WebUI handlers, agent profiles, prompts, skills, projects, runtime boundaries, and contribution workflow. Load the focused reference files before giving implementation guidance.
brainstorming
Socratic questioning protocol + user communication.
himalaya
Read, search, and organize emails via the himalaya CLI (IMAP/SMTP). Use when the user asks to check email, search messages, list folders, download attachments, or manage email flags and folders.
cron
Create and manage scheduled tasks that run automatically — periodic literature checks, data collection, report generation, and recurring notifications. Use when the user wants to schedule, list, pause, resume, or delete cron jobs via researchclaw cron.
convex-agents
Building AI agents with the Convex Agent component including thread management, tool integration, streaming responses, RAG patterns, and workflow orchestration.
orbit-notion
Open Orbit briefing skill — selected by the Orbit pipeline when Notion is the user's only connected connector, or when the user explicitly scopes their daily digest to Notion. Pulls the past 24 hours of document edits, comments, mentions, and database row changes from the user's authenticated Notion connection and…