← 返回 Skills 市场
oiioai

Emotional Safety Fuse

作者 OiiOAI · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
240
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install emotional-safety-fuse
功能描述
Add an emotional safety layer to reminiscence, life review, grief, trauma-adjacent, or otherwise emotionally loaded conversations by detecting distress signa...
使用说明 (SKILL.md)

emotional-safety-fuse

Safety protocols for conversations involving emotional depth, reminiscence, or sensitive topics. Detects distress signals and provides de-escalation strategies.

Description

A plug-and-play safety layer for any agent that handles emotionally loaded conversations. It defines distress signals, immediate response steps, grounding options, recovery rules, and hard boundaries.

When to Use

  • Conversations involving personal loss, trauma, grief, illness, or conflict
  • Life review or reminiscence sessions
  • Deep autobiographical or identity-focused disclosure
  • Any conversation where the person becomes visibly strained, withdrawn, or overwhelmed

Distress Signals

Treat any of the following as a cue to slow down or stop probing:

  • The person says they do not want to continue
  • Crying, long silence, strained voice, repeated sighing, or sudden agitation
  • Short, shrinking answers after a previously open conversation
  • Defensive, avoidant, or abruptly detached responses
  • Repeating "it's nothing," "never mind," or similar deflections while sounding distressed
  • Noticeable confusion, disorientation, or inability to stay with the thread
  • A sudden shift after mentioning bereavement, violence, displacement, abuse, or other potentially traumatic material

Response Protocol

When distress appears, use this sequence:

  1. Stop: End all deepening questions immediately.
  2. Validate: Acknowledge the reality of the moment without analyzing it.
  3. Ground: Help the person orient to the present using one simple prompt.
  4. Offer choices: Give control back. Continue, pause, switch topics, or end.

Examples:

  • "We do not have to keep going with this part."
  • "That sounds like a very weighty memory."
  • "Would it help to pause and notice what is around you right now?"
  • "We can rest, change topics, or stop here. You can choose."

Grounding Options

Use only one grounding prompt at a time.

  • Ask about the current room, weather, light, or a nearby object
  • Suggest a sip of water or a brief pause
  • Invite the person to notice their feet, chair, or breathing without making it clinical
  • Shift attention to something neutral in the present environment

Avoid stacking choices or turning grounding into a performance task.

Recovery Strategy

  • After de-escalation, stay with gentle, low-demand conversation for at least two turns.
  • Re-enter depth only if the person clearly leads the way back.
  • If resuming, return through something concrete and low-intensity.
  • If distress returns, de-escalate again and do not push for completion.
  • Story quality, memory completeness, and productivity goals are always secondary to safety.

Red Lines

  • Do not continue probing after the person says no
  • Do not argue with, reinterpret, or minimize their reaction
  • Do not pressure them to find meaning while distressed
  • Do not use guilt, urgency, or therapeutic language you are not qualified to support
  • Do not promise safety, healing, confidentiality, or clinical benefit
  • Do not escalate the conversation for the sake of richer output

Escalation Boundary

This skill supports conversational de-escalation, not crisis intervention or therapy. If there are signs of immediate danger, self-harm risk, medical emergency, abuse requiring urgent reporting, or severe disorientation, switch from reminiscence support to the host system's emergency or professional-help protocols.

安全使用建议
This skill is internally coherent and lightweight, but before installing, verify the host agent's concrete behaviors: 1) Map the SKILL.md's 'host system emergency/professional-help protocols' to explicit, tested actions so the agent doesn't attempt inappropriate escalation. 2) Decide and document how sensitive conversational content is logged, stored, or transmitted (avoid retaining more than necessary and ensure consent). 3) Confirm how the skill is invoked and whether automated interventions (pausing or suggesting resources) are acceptable for your users. 4) Test with domain experts (clinicians or trained facilitators) to validate phrasing, cultural and accessibility adaptations, and thresholds for distress detection to minimize false positives/negatives. 5) Do not rely on this skill as a substitute for crisis services—ensure a clear and tested path to professional emergency support when indicated.
功能分析
Type: OpenClaw Skill Name: emotional-safety-fuse Version: 1.0.0 The skill bundle 'emotional-safety-fuse' (SKILL.md) provides behavioral guidelines for an AI agent to handle emotionally sensitive topics like grief or trauma with empathy and safety. It contains no executable code, network requests, or instructions for data exfiltration, focusing entirely on conversational de-escalation and user well-being.
能力评估
Purpose & Capability
Name and description (emotional-safety/de-escalation) match the SKILL.md content: defined distress signals, stop/validate/ground/offer-choices flow, recovery and escalation boundaries. No unrelated env vars, binaries, or install steps are requested.
Instruction Scope
Instructions stay within conversational safety boundaries and do not ask to read files, access credentials, or call external endpoints. The SKILL.md references switching to the host system's emergency or professional-help protocols without defining them — this is reasonable but somewhat open-ended and should be mapped to concrete host behaviors before deployment.
Install Mechanism
No install spec and no code files (instruction-only). Lowest-risk delivery: nothing is written to disk or downloaded by the skill itself.
Credentials
The skill requires no environment variables, credentials, or config paths. Its needs are minimal and appropriate for an instruction-only safety layer.
Persistence & Privilege
always is false and the skill does not request system-wide persistence or modify other skills. Model invocation is allowed (platform default) which is appropriate for a conversational safety layer; no additional privileged settings are present.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install emotional-safety-fuse
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /emotional-safety-fuse 触发
  4. 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
v1.0.0
- Initial release of emotional-safety-fuse. - Adds a safety layer for emotionally loaded conversations by detecting distress signals. - Defines immediate response steps: pause depth work, validate, ground the person, and offer choices. - Provides clear distress signals, grounding options, and red lines for safe conversational handling. - Designed for reminiscence, life review, grief, trauma-adjacent, and other sensitive discussions. - Calls for escalation to emergency protocols in cases of immediate danger or severe distress.
元数据
Slug emotional-safety-fuse
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Emotional Safety Fuse 是什么?

Add an emotional safety layer to reminiscence, life review, grief, trauma-adjacent, or otherwise emotionally loaded conversations by detecting distress signa... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 240 次。

如何安装 Emotional Safety Fuse?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install emotional-safety-fuse」即可一键安装,无需额外配置。

Emotional Safety Fuse 是免费的吗?

是的,Emotional Safety Fuse 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Emotional Safety Fuse 支持哪些平台?

Emotional Safety Fuse 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Emotional Safety Fuse?

由 OiiOAI(@oiioai)开发并维护,当前版本 v1.0.0。

💬 留言讨论