← 返回 Skills 市场
1273
总下载
0
收藏
3
当前安装
4
版本数
在 OpenClaw 中安装
/install principle-synthesizer
功能描述
Synthesize invariant principles from 3+ sources — find the core that survives across all expressions.
安全使用建议
This is an instruction-only skill whose behavior is coherent with its description. Before installing or using it: (1) avoid submitting highly sensitive data if your agent uses a cloud LLM (the SKILL.md explicitly says processing uses the agent's configured model); (2) ensure you provide at least 3 independent sources as required; (3) review any Golden Master candidates manually — the skill produces candidates, not guaranteed truths; and (4) if you plan to feed outputs from other skills, confirm those skills' outputs are trustworthy. No install or credential risks were found in the skill itself.
功能分析
Type: OpenClaw Skill
Name: principle-synthesizer
Version: 1.0.3
The 'Principle Synthesizer' skill bundle consists of metadata and markdown instructions (SKILL.md) designed to guide an AI agent in distilling invariant principles from multiple text sources. The logic focuses entirely on data processing, normalization, and classification (N≥3 validation) without any executable code, shell commands, or network requests. There are no indicators of prompt injection attacks, data exfiltration, or unauthorized system access; the skill explicitly states it operates within the agent's trust boundary and does not write to disk.
能力评估
Purpose & Capability
Name/description (synthesizing invariant principles from multiple sources) match the SKILL.md: all functionality is text normalization, alignment, and summarization. The skill does not request unrelated binaries, credentials, or config paths and only references other complementary skills (pbe-extractor, essence-distiller, principle-comparator) as possible input sources, which is reasonable.
Instruction Scope
SKILL.md contains a clear methodology (gather → align → validate → classify → output) and explicit boundaries (N≥3, human judgment required). It does not instruct the agent to read unrelated system files, use env vars, or call third-party endpoints. Note: the SKILL.md states that processing uses the agent's configured model — so if the agent is configured to use a cloud-hosted LLM, user data will be processed by that external service (this is expected but worth awareness).
Install Mechanism
No install spec and no code files (instruction-only). Nothing is written to disk or downloaded by the skill itself, so install risk is minimal.
Credentials
The skill declares no required environment variables, credentials, or config paths. It does reference inputs from other skills but does not request access to unrelated secrets. This is proportionate to a text-synthesis task. Reminder: the agent's model credentials/config are outside the skill and determine where data is sent.
Persistence & Privilege
always is false and the skill is user-invocable; it does not request persistent system presence, nor does it modify other skills' configurations. Autonomous invocation is allowed by platform default but not elevated by this skill.
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 principle-synthesizer - 安装完成后，直接呼叫该 Skill 的名称或使用
/principle-synthesizer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
No user-visible changes in this version.
- No file changes detected between versions 1.0.2 and 1.0.3.
v1.0.2
- Added explicit data handling section: clarifies that all synthesis is local to your agent, with no external API calls or disk writes by the skill.
- Updated tags to better reflect synthesis, merging, and knowledge-management focus.
- Revised homepage URL and minor metadata.
- No algorithmic or functional changes to synthesis logic.
v1.0.1
Migrated to public GitHub repo, updated homepage URLs
v1.0.0
Initial release of Principle Synthesizer: synthesize invariant principles from 3+ sources (Golden Master candidates).
- Identifies core principles that survive semantic alignment across 3 or more sources (N≥3* validation).
- Distinguishes between invariant, domain-specific, and noise principles with classification and evidence.
- Outputs structured synthesis results, metrics, and candidates for canonical documentation.
- Includes user-invocable prompts, error handling, and concise summary (share_text) when genuine Golden Master candidates are found.
- Designed for rigorous, transparent multi-source distillation and methodology creation.
元数据
常见问题
Principle Synthesizer 是什么？
Synthesize invariant principles from 3+ sources — find the core that survives across all expressions. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 1273 次。
如何安装 Principle Synthesizer？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install principle-synthesizer」即可一键安装，无需额外配置。
Principle Synthesizer 是免费的吗？
是的，Principle Synthesizer 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Principle Synthesizer 支持哪些平台？
Principle Synthesizer 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 Principle Synthesizer？
由 Lee Brown（@leegitw）开发并维护，当前版本 v1.0.3。
推荐 Skills