Flux7 Mesh — AI skill for Claude Code
Guardrail sidecar proxy for AI agents — policy, approval gates, grants, tracing.
How to install Flux7 Mesh
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open KTCrisis/flux7-mesh and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Flux7 Mesh does
Guardrail sidecar proxy for AI agents — policy, approval gates, grants, tracing. One Go binary, one YAML.
Alternatives in AI
- Axonhub — ⚡️ Open-source AI Gateway — Use any SDK to call 100+ LLMs 3.2k ★
- Adopt AI Properly — Workflow recipe — the org-side AI adoption arc, policy to proof, by chaining 4 skills 1.3k ★
- Cupcake — A native policy enforcement layer for AI coding agents 255 ★
README
______ ____ __
/ __/ //_ /_____ _ ___ ___ / /
/ _// /__/ /___/ ' \/ -_|_- _ \
/_/ /____/_/ /_/_/_/\__/___/_//_/
flux7-mesh
[](https://github.com/KTCrisis/flux7-mesh/releases) [](https://github.com/KTCrisis/flux7-mesh/actions/workflows/ci.yaml) [](https://go.dev) [](LICENSE)
**Guardrail for AI agents.** Open-source sidecar proxy between AI agents and their tools — policy, human approval, and tracing without changing agent code.
One binary. One YAML config. Fail closed by default.
Works with Claude Code, Cursor, Anthropic Managed Agents, LangChain, CrewAI, or any agent that uses HTTP, MCP, or CLI tools.
Table of contents
- Architecture
- Install
- Quick start
- Features — approval, grants, rate limiting, tracing, OTEL, supervisor
- Documentation
- Roadmap
Architecture
flowchart LR
subgraph Agents["Agents"]
A1["Claude Code / Cursor"]
A2["LangChain / CrewAI"]
A3["Any HTTP agent"]
end
subgraph Mesh["flux7-mesh (sidecar proxy)"]
direction TB
REG["Registry
(tools)"]
RL["Rate limiter
+ loop detect"]
POL["Policy engine
(glob, conditions)"]
FWD["Forward"]
APP["Approval store"]
GRT["Grant store
(sudo for agents)"]
TRC["Trace store
(JSONL + sessions)"]
OTEL["OTEL exporter
(file / stdout / OTLP)"]
REG --> RL --> POL --> FWD
POL -.approval.-> APP
POL -.bypass.-> GRT
FWD --
Related Skills
Flux7 Supervisor
Standalone L1 supervisor agent for flux7-mesh — rules engine + pluggable LLM (Ollama, Anthropic, Claude Code M
Agentic Review Gate
Governed, auditable AI review platform: the review and approval work a team does by hand, as a webhook-trigger
Salient Core
Guardrail and permission layer for AI agent harnesses. A default-deny control kernel that gates every tool cal
Yea
HTTP was built for browsers. YEA is built for agents. YEA (Your Explicit Approval) is an open protocol for AI
Asynq Agentd
An open-source daemon that keeps your AI coding agents (Claude Code, Codex, OpenCode) running 24/7 on any alwa
Always On Agent
Agent Skill for building a personal AI agent that runs continuously on your own machine: four layers to keep s
Related Agents
Gateway Notes
Purpose: the confirmed gateway/guardrail mechanisms for the workshop, input guard, output sidecar, MCP authz,
Infra Engineer.Agent
Use for repository-grounded IaC changes and scoped deployment work. Plan-only by default; apply/rollback is al
Compile-time Profiling
clang++ -ftime-trace -ftime-trace-granularity=1 file.cpp chrome://tracing # Load the JSON bloaty binary -d sym