← 返回 Skills 市场
Training Report
作者
Samuel Berthe
· GitHub ↗
· v1.0.0
· MIT-0
64
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install training-report
功能描述
Produce a professional training/workshop report as a .docx file. Use this skill whenever the user mentions "training report", "workshop report", "compte rend...
安全使用建议
This skill appears to be a coherent, instruction-only writer for training/workshop reports, but take three precautions before installing or enabling it:
1) Trigger sensitivity: the skill's description and SKILL.md encourage activation on many natural-language 'training' phrases. If you also use the agent for non-document 'training' tasks (e.g., ML pipelines), the skill may misactivate and begin an interview workflow. Ask the platform owner to narrow or review the activation triggers or to make the skill user-invocable so it runs only when explicitly requested.
2) Data sensitivity: the interview asks for participant names, behaviors, and contact details. Only provide PII you are permitted to share. Prefer anonymized/group-level feedback where appropriate and confirm any named negative feedback is acceptable to include before uploading or entering it.
3) Docx dependency and file handling: the skill will try to load a separate 'docx' skill to build the .docx and may ask you to upload a template file. Confirm how your platform handles file unpacking/repacking and where output files will be written. Because this skill delegates .docx work to another skill, review that docx skill's permissions before proceeding.
If you want lower risk: keep this skill disabled for autonomous invocation and call it manually, or ask the maintainer to refine triggers (e.g., require explicit 'write a training report' rather than any occurrence of 'training').
功能分析
Type: OpenClaw Skill
Name: training-report
Version: 1.0.0
The 'training-report' skill bundle is a well-structured tool designed to help an AI agent conduct interviews and generate professional workshop reports in Markdown and Word formats. The instructions in SKILL.md and the reference files (tone-of-voice.md, markdown-draft.md, docx-generation.md) provide detailed, legitimate guidance on document structure, linguistic nuances, and diplomatic reporting. There is no evidence of malicious intent, data exfiltration, or prompt-injection attacks; the skill uses standard OpenClaw tools for file management and user interaction as expected for its stated purpose.
能力评估
Purpose & Capability
The name and description match the instructions: this is an instruction-only skill to draft training/workshop reports and produce a .docx. There are no unrelated environment variables, binaries, or installs declared. One mismatch to note: the SKILL.md's trigger guidance is very broad (many natural-language triggers like 'training') which risks activation for non-document tasks that use the word 'training' (the evals explicitly test for an ML training pipeline trap). Functionality requested (asking for session details, templates, generating .md/.docx) is proportionate to the stated purpose.
Instruction Scope
The runtime instructions are detailed and prescriptive (structured interview, step-by-step draft-and-approve flow, docx-template unpack/inject/repack). That scope is appropriate for generating reports, but the SKILL.md repeatedly instructs to 'Always use this skill' for many triggers which is too permissive. The risk: the skill's workflow could be applied in contexts where the user's intent is not to write a report (e.g., ML model training), causing unnecessary or confusing questions and potential disclosure of participant-sensitive information. The instructions ask for participant names, behaviors, and contact details (expected for reports) — these are sensitive PII and the skill does not include explicit safeguards beyond diplomatic guidance. The skill also instructs the agent to seek and use other skills (docx, humanizer) which is expected but broadens its runtime surface.
Install Mechanism
This is instruction-only with no install spec, no downloaded code, and no external package installs — lowest risk for install-time compromise.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. It does request user-supplied content (templates, participant feedback) which is reasonable for its purpose, but users should treat that content as potentially sensitive.
Persistence & Privilege
The skill is not 'always: true' and does not request elevated persistence. However, it is marked non-user-invocable (user-invocable: false) while allowing autonomous invocation by the agent (disable-model-invocation: false). Combined with the broad trigger guidance in the description, this increases the risk of unexpected autonomous activation. Autonomous invocation alone is normal, but here it amplifies the over-trigger concern.
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 training-report - 安装完成后，直接呼叫该 Skill 的名称或使用
/training-report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
v1.0.0
training-report v1.0.0 — Initial Release
- Generates professional training or workshop reports in .docx format, with a robust, step-by-step workflow.
- Triggers automatically for any request to document training, workshops, or onboarding events (in English or French).
- Guides user through structured interview batches to gather metadata, context, session walkthrough, participant feedback, recommendations, and annexes.
- Supports brand customization through Word template uploads or simple color branding.
- Checks for essential dependencies (docx skill) and recommends humanizer skill for polished final drafts.
- Produces a Markdown draft first, ensuring transparency and quality before .docx export.
元数据
常见问题
Training Report 是什么？
Produce a professional training/workshop report as a .docx file. Use this skill whenever the user mentions "training report", "workshop report", "compte rend... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 64 次。
如何安装 Training Report？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install training-report」即可一键安装，无需额外配置。
Training Report 是免费的吗？
是的，Training Report 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Training Report 支持哪些平台？
Training Report 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 Training Report？
由 Samuel Berthe（@samber）开发并维护，当前版本 v1.0.0。
推荐 Skills