← 返回
未分类 中文

OMNI Semantic Signal Engine

Local-only semantic context filtering that saves up to 90% in token costs.
本地化语义上下文过滤,可节省高达90%的Token费用
fajarhide
未分类 clawhub v0.5.9 1 版本 100000 Key: 无需
★ 0
Stars
📥 355
下载
💾 0
安装
1
版本
#latest

概述

OMNI Semantic Signal Engine

Less noise. More signal. Cut your AI token consumption by up to 90%.

This skill provides a secure proxy to OMNI's powerful semantic distillation engine. It facilitates the filtering of terminal output from noisy commands, ensuring your agent focuses on critical information.

> [!IMPORTANT]

> Dependency Required: This skill is a proxy and requires the OMNI CLI binary to be installed on your local system path.

Tools Included

omni_cmd

Execute terminal commands and development tools (npm, git, cargo, etc.) through OMNI. Use this for all your standard development tasks to keep your context window lean and your API bills low.

omni_rewind

Retrieve full archived logs if OMNI's distilled summary was too aggressive.

Security & Trust (Design Intent)

This skill is designed as a secure bridge to the local OMNI binary. To maintain transparency regarding its security properties:

  • Node-Level Sanitization: The plugin explicitly strips ~25 dangerous environment variables at the process level before execution (e.g., LD_PRELOAD, NODE_OPTIONS, BASH_ENV).
  • Safe Process Spawning: Uses execFile with an atomic argument array, bypassing shell string parsing and reducing injection risks.
  • Local-Only Architecture: OMNI is designed as a local-first engine. No terminal data or telemetry is exfiltrated by the engine during normal operation.
  • Data Persistence: Archives for the omni_rewind feature are stored strictly on your local device in a SQLite database (~/.omni/omni.db).
  • Audit Invitation: Both the plugin code and the OMNI binary source are open-source. We encourage security-conscious users to audit the code at the OMNI GitHub Repository.

版本历史

共 1 个版本

  • v0.5.9 当前
    2026-05-07 14:17 安全 安全

安全检测

腾讯云安全 (Keen)

安全,无风险
查看报告

腾讯云安全 (Sanbu)

安全,无风险
查看报告

🔗 相关推荐

ai-agent

Self-Improving + Proactive Agent

ivangdavila
自我反思+自我批评+自我学习+自组织记忆。智能体评估自身工作、发现错误并持续改进。
★ 1,434 📥 327,507
ai-agent

self-improving agent

pskoett
记录自身发现以实现自我改进的技能
★ 4,153 📥 925,648
ai-agent

Find Skills

root
帮助用户发现和安装智能体技能,当用户询问如「如何做X」、「找X的技能」、「有能做...的吗」等问题时
★ 1,507 📥 567,317