Jig-ToolGuard is the first pre-execution agent firewall that intercepts tool calls at the code level before they execute, enforcing hard constraints beyond prompt-level guardrails.

Stars

115

7-day growth

No data

Forks

17

Open issues

0

License

MIT

Last updated

2026-07-27

AI repository intelligence
FR-AI / ANALYSIS

Why it is worth attention

It uniquely provides code-level interception with whitelist/denylist enforcement, unlike all major agent frameworks that rely on prompt suggestions, and includes 11 preset agents, loop engineering, and 4-layer memory.

Who it is for

  • Developers building secure multi-agent systems
  • AI safety engineers needing runtime enforcement
  • Teams using DeepSeek models for automation
  • Researchers evaluating agent guardrail approaches

Use cases

  • Securing automated code review and execution pipelines
  • Running multi-agent development workflows with safety gates
  • Building safe LLM-powered DevOps or security audit agents
  • Enforcing hard constraints on tool calls in production AI applications

Strengths

  • Pre-execution interception at code level, not prompt-level
  • Three-layer safety (whitelist, denylist, runtime hook) for hard constraints
  • 11 pre-built agents covering PM, coding, review, security, DevOps, etc.
  • Loop engineering with convergence detection, quality validation, and checkpoint resume

Considerations

  • Alpha release (v0.6.3) with features like Meta-Harness still in progress
  • Requires Python 3.10+ and a DeepSeek API key; only DeepSeek and OpenAI providers documented
  • No explicit support for non-OpenAI/DeepSeek models or production-grade durability yet

README quick start

Quick Start

Description

Tree-SOP Agent — 群聊式多 Agent 软件开发框架。预置 11 个 Agent 角色,完整 SOP 管道,Harness 约束体系,4 层记忆架构。

Related repositories

Similar projects matched by category, topics, and programming language.

lopopolo
Featured
lopopolo GitHub avatar

harness-engineering

Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.

AI & Machine LearningAI Agents
2,390
slvDev
Featured
slvDev GitHub avatar

esp32-ai

A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.

AI & Machine LearningLarge Language Models
1,960
littledivy
Featured
littledivy GitHub avatar

mimic

mimic captures traffic from any iOS or web app and automatically generates a Python client library that lets you call the app's API like a regular library.

AI & Machine Learning
1,482