← 返回 Skills 市场

Flow State Monitoring

Name: Flow State Monitoring
Author: zvirb

作者 zvirb · GitHub ↗ · v1.0.3 · MIT-0

cross-platform ⚠ suspicious

112

总下载

当前安装

版本数

在 OpenClaw 中安装

/install flow-state-monitoring

功能描述

Workflow-driven skill that infers deep focus and autonomously mutes interruptions.

使用说明 (SKILL.md)

Lean Philosophy (Principles)\r

Kaizen (改善): This skill is an atomic node, broken down into its simplest, smallest component to eliminate waste and ensure perfection.\r
Standardized Work (Hyojun Sagyo): This node represents the most efficient, standardized path for this specific task before automation.\r
Jidoka (自働化): This node includes autonomous defect detection. It will stop immediately and report if it cannot achieve the expected outcome.\r \r

Flow State Monitoring\r

\r This skill orchestrates a workflow to analyze focus telemetry and autonomously update your status in Google Workspace to protect your flow state.\r \r

Cognitive Directives\r

WHEN [telemetry from local sensors indicates deep focus] \r THEN [\r Execute the following Jidoka-validated loop:\r

Execute llm_classify_intent (Sub-Agent) to classify the activity.\r
- Verification Step (Jidoka): Verify the sub-agent returns a valid classification (e.g., "Deep Work"). IF it hallucinates text, retry classification.\r
IF classified as "Deep Work", Execute the native terminal command gog calendar create primary --summary "Busy - In Flow" to block interruptions.\r
- Verification Step (Jidoka): Check if the calendar event creation returns a successful JSON response. IF it fails, wait 3 seconds and retry (max 3 times). IF it still fails, report the error and STOP.\r ]\r \r

Expected Output\r

A JSON confirmation of the calendar status update.\r

安全使用建议

Before installing, verify what 'gog' is and where it comes from (official Google tool, third-party, etc.). Ask the skill author to: (1) declare required credentials or config paths (where 'gog' reads OAuth tokens), (2) describe exactly which local sensors or files will be read for telemetry, and (3) add an explicit user confirmation step before creating calendar events or provide an opt-in/opt-out. If you decide to test it, run it with a disposable/low-privilege Google account and inspect the 'gog' configuration files (and any token caches) so you know what will be used. If you cannot get these clarifications, treat the skill as risky because it can autonomously modify your calendar and may access local telemetry without clear limits.

功能分析

Type: OpenClaw Skill Name: flow-state-monitoring Version: 1.0.3 The skill defines a workflow for automating Google Calendar status updates based on focus telemetry using the 'gog' CLI tool. The instructions in SKILL.md are well-structured, including error-handling logic (Jidoka) and clear intent classification, with no evidence of malicious behavior, data exfiltration, or unauthorized system access.

能力评估

ℹ Purpose & Capability

The skill claims to update Google Workspace calendar status and the SKILL.md runs a calendar-creation command via the 'gog' binary, which is coherent. However, the skill does not declare any required credentials or config paths for Google access (OAuth tokens/config), relying on an assumption that the 'gog' CLI is pre-authenticated. The source/homepage are unknown, which increases uncertainty about that binary's behavior.

⚠ Instruction Scope

Instructions refer to "telemetry from local sensors" but do not define what sensors, what data sources, or what commands to run to obtain telemetry; that gives the agent wide latitude to access local system data. It also directs an LLM sub-agent (llm_classify_intent) and the native command `gog calendar create ...` to modify the user's calendar; the destination for error reporting is unspecified. The lack of concrete telemetry access constraints is scope creep and a privacy risk.

✓ Install Mechanism

This is instruction-only (no install spec, no code files), so nothing new is written to disk by the skill itself. It does require the 'gog' binary to be present on PATH; the skill doesn't provide or vet that binary or its origin.

⚠ Credentials

The skill lists no required environment variables or config paths but performs calendar writes. That implies it expects an existing authenticated CLI configuration (OAuth tokens or similar) to be present, which is not declared. Because the skill can modify calendar state, it should explicitly declare required credentials/config locations and justify them; the omission is disproportionate to the transparency a user needs.

ℹ Persistence & Privilege

always is false (good). The skill allows autonomous model invocation (default), which means the agent could autonomously create calendar events based on telemetry. Autonomous invocation combined with unclear telemetry sources and undeclared credential use increases the potential impact, though autonomous invocation alone is expected behavior.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install flow-state-monitoring
安装完成后，直接呼叫该 Skill 的名称或使用 /flow-state-monitoring 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.3

- Improved workflow to include Jidoka-validated loop for greater reliability. - Added verification steps: sub-agent output and calendar event creation are now checked for validity. - Implemented retry logic for failed classifications and event creation (up to 3 attempts). - Enhanced error handling: failures are reported and process stops if retries are exhausted.

v1.0.2

flow-state-monitoring 1.0.2 - Added "gog" to the required binaries list. - Updated directive to use the native terminal command: gog calendar create primary --summary "Busy - In Flow". - Clarified cognitive workflow for calendar status updates.

v1.0.1

- Expanded OS support from "windows" to "all". - No other changes to functionality or workflow.

v1.0.0

- Initial release of the flow-state-monitoring skill. - Monitors local focus telemetry to detect deep work states. - Automatically updates Google Workspace calendar status to "Busy - In Flow," muting potential interruptions. - Follows Lean principles: atomic design, standardized workflow, and built-in defect detection with automated error reporting.

元数据

Slug flow-state-monitoring

版本 1.0.3

许可证 MIT-0

累计安装 1

当前安装数 1

历史版本数 4

常见问题