siegard-code

agent
Guvenlik Denetimi
Uyari
Health Uyari
  • License — License: Apache-2.0
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 6 GitHub stars
Code Uyari
  • Code scan incomplete — No supported source files were scanned during light audit
Permissions Gecti
  • Permissions — No dangerous permissions requested
Purpose
This project is an autonomous agent infrastructure for Claude Code that manages the entire software development lifecycle. It coordinates 18 specialized agents across three teams to automatically write specifications, plan backlogs, implement features, run QA, and deliver tested code.

Security Assessment
The automated audit found no hardcoded secrets or dangerous permissions. However, the static code scanner was unable to analyze the underlying source files, meaning a manual security review of the code is not possible through automated tooling alone. As an AI-driven development tool, it inherently requires permissions to execute shell commands, read/write local files, and potentially make network requests to function properly. Without verifiable source code analysis, developers must assume the standard risks associated with giving an autonomous agent read and write access to their local environment. Overall risk: Medium.

Quality Assessment
The repository is quite new and lacks community visibility, currently sitting at only 6 GitHub stars. On a positive note, it was recently updated (pushed 0 days ago), indicating active development, and it clearly defines its terms under the highly permissive Apache-2.0 license. The comprehensive documentation outlines a complex architecture, but the extremely low community adoption means there has been minimal third-party testing or public vetting. Community trust level is currently low.

Verdict
Use with caution: the tool is active and well-documented, but extremely low community adoption and a lack of verifiable source code scanning mean you should manually inspect its behavior before granting it autonomous access to your environment.
SUMMARY

Most AI coding tools help you write code. Siegard Code manages the entire development lifecycle — it writes specifications, plans backlogs, implements features, runs QA, and delivers tested code. All autonomously, all traceable, all through Claude Code.

README.md

Siegard Code

Siegard Code

Autonomous agent infrastructure for Claude Code that ships features — from spec to delivery.

Version License Works with Claude Code 18 Agents 20+ Skills

18 specialized agents. 3 coordinated teams. One command to orchestrate them all.


Most AI coding tools help you write code. Siegard Code manages the entire development lifecycle — it writes specifications, plans backlogs, implements features, runs QA, and delivers tested code. All autonomously, all traceable, all through Claude Code.

You describe what you want. Siegard's agents handle the rest: the Spec team writes and validates technical specifications, the Dev team plans and implements Stories with automated QA, and the Reverse Spec team can even document existing codebases. Every step produces artifacts. Every decision is logged. Every quality gate has teeth.

This is not a product. It is the agent infrastructure you install into your own projects.


Why Siegard Code?

Pain point How Siegard solves it
"AI writes code but ignores the spec" Specs are the single source of truth — agents trace every line back to a Use Case
"No one documents anything" Documentation is a byproduct, not a chore — specs, backlogs, QA reports generated automatically
"Context gets lost between sessions" Session resume protocol reconstructs state from disk artifacts — pick up where you left off
"AI makes changes I didn't ask for" Human confirmation required at every major gate — nothing ships without your approval
"Testing is always an afterthought" QA is built into the pipeline — every Story passes through automated test validation

Architecture

Three agent teams work in sequence, connected by formal handoff protocols:

graph TB
    subgraph INPUT["Input"]
        REQ[Requirement / Existing code]
    end

    subgraph RSPEC["Reverse Spec Team · 3 agents"]
        RSORC[Orchestrator]
        RSANA[Analyzer]
        RSWRT[Writer]
        RSORC --> RSANA --> RSWRT
    end

    subgraph SPEC["Spec Team · 6 agents"]
        SPORC[Orchestrator]
        SPWRT[Spec Writer]
        SPREV[Spec Reviewer]
        SPBCK[Back Spec Agent]
        SPFRT[Front Spec Agent]
        SPVAL[Spec Validator]
        SPORC --> SPWRT --> SPREV
        SPREV -->|APPROVED| SPBCK
        SPREV -->|REJECTED| SPWRT
        SPBCK --> SPVAL
        SPVAL -->|all back.md valid| SPFRT
        SPFRT --> SPVAL
        SPVAL -->|INVALID| SPBCK
        SPVAL -->|INVALID front| SPFRT
    end

    subgraph DEV["Dev Team · 9 agents"]
        DVORC[Orchestrator-Dev]
        PLAN[Planner]
        UIAG[UI Agent]
        DEVAG[Developer]
        QAAG[QA & Docs]
        DVORC --> PLAN --> UIAG --> DEVAG --> QAAG
        QAAG -->|REJECTED| DEVAG
    end

    REQ --> RSORC
    REQ --> SPORC
    RSWRT -->|specs · draft| SPORC
    SPVAL -->|VALID| DVORC
    QAAG -->|APPROVED| FIN[Delivery]
Team Agents What it does
Spec Orchestrator, Writer, Reviewer, Back Spec, Front Spec, Validator Writes, reviews, and validates technical specifications
Dev Orchestrator, Planner, UI Agent, Developer, QA & Docs (×2 FE/BE) Plans backlogs, implements code, runs QA
Reverse Spec Orchestrator, Analyzer, Writer Reverse-engineers specs from existing code

Quick Start

Prerequisites

Install into your project

Copy the dist/.claude/ directory into your project:

cp -r dev-team/dist/.claude/ /path/to/your-project/.claude/

This copies all agents, skills, and commands into your project's .claude/ directory.

Configure your project

Add these fields to your project's CLAUDE.md:

domain: backend           # or "frontend" — determines which pipeline runs
stack: Node.js, Express   # your tech stack (agents adapt, not hardcode)
specs_dir: docs/specs     # where specifications live

Run your first command

# Create a specification from requirements
/u-spec docs/specs

# Implement from approved specs
/u-dev docs/specs my-feature

That's it. The orchestrator detects the mode, activates the right agents, and guides you through every step.


Commands

Six slash commands cover the entire development lifecycle:

Command Purpose Output
/u-spec Create or evolve technical specifications Approved specs in {SPECS_DIR}
/u-dev Run a development session (plan → implement → QA) Tested, delivered code
/u-reverse-spec Generate specs from existing source code Draft specs for review
/u-spec-triage Fix spec validation errors incrementally Corrected specs
/u-improve Capture an improvement request improve##.md
/u-bug-report Capture a structured bug report bug##.md

Workflows

Feature (full):       /u-spec  →  /u-dev
Quick improvement:    /u-improve  →  /u-dev
Bug fix:              /u-bug-report  →  /u-dev
Reverse + evolve:     /u-reverse-spec  →  /u-spec  →  /u-dev

Key Features

Spec-First Development

Every feature starts with a validated specification — openapi.yaml, use cases, business rules, state machines, screen specs, and navigation flows. No code is written without an approved spec.

Automatic Mode Detection

The orchestrator reads your project state and selects the right mode: Spec-first, Improve, Bug, Resume, or Reverse-eng review. No flags, no configuration — it just works.

Frontend & Backend Pipelines

One command, two pipelines. The domain: field in your CLAUDE.md routes to the correct agents. Frontend gets a UI Agent for visual specs; backend gets infrastructure dependency tracking.

Quality Gates with Escalation

  • Spec Reviewer: Max 3 rejection cycles before human escalation
  • Spec Validator: Cross-reference validation with coverage reports
  • QA Agent: Test-gate + full mode with max 3 rework rounds
  • Nothing passes silently — every gate produces artifacts

Token-Efficient Context Management

  • Context mounting: Each agent receives only what it needs
  • Short mode: Reactivations use ~2K tokens instead of ~15K
  • Triage mode: Process 5–10 errors per session, not 20+ at once

Session Resilience

Interrupted? Run the same command again. The orchestrator reads its log file, reconstructs state from disk artifacts, and resumes from the exact stopping point.


Project Structure

your-project/
├── .claude/
│   ├── agents/
│   │   ├── dev/                  # Dev team (FE + BE agents + protocols)
│   │   ├── spec/                 # Spec team (6 agents + protocols)
│   │   └── reverse-spec/         # Reverse Spec team (3 agents + protocols)
│   ├── commands/                 # 6 slash commands
│   └── skills/                   # 20+ reusable skill libraries
├── CLAUDE.md                     # Your project configuration
├── {SPECS_DIR}/
│   ├── _global/                  # Shared: conventions, error codes, glossary
│   ├── domains/{domain}/         # Backend: openapi.yaml + spec.md + back.md
│   └── front/                    # Frontend: screens, flows, design system
└── {SESSIONS_DIR}/
    └── {SESSION}/                # Current session working directory
        ├── backlog.md            # Planned Stories
        ├── us-XX-delivery.md     # Delivery artifacts
        ├── us-XX-qa.md           # QA reports
        └── log-orchestrator-dev.md

Action Guide

I want to... Commands
Build a complete feature /u-spec/u-dev
Implement from existing specs /u-dev
Fix a bug /u-bug-report/u-dev
Apply a quick improvement /u-improve/u-dev
Document existing code /u-reverse-spec/u-spec
Fix spec validation errors /u-spec-triage
Report a spec problem from dev /u-spec (reverse feedback mode)

Documentation

Section Description
Overview Architecture, concepts, glossary
Installation Setup and project configuration
Commands All 6 commands in detail
Agent Teams Spec, Dev, and Reverse Spec teams
Execution Flows Step-by-step workflows
Protocols 15 on-demand behavioral protocols
Skills 20+ reusable skill libraries
Artifacts Generated files and lifecycle
Estimates Token and time budgets
Resilience Failure scenarios and recovery
Reference Cheat sheet and action guide

Estimates

Mode Tokens Time Scope
New spec (full pipeline) ~19K 7–12 min Per domain
Fast-track spec change ~11K 4–7 min Per change
Development (spec-first) ~14K 10–18 min Per Story
Development (improve) ~10K 8–13 min Per Story
Triage ~3K 1–2 min Per item

License

This project is licensed under the Apache License 2.0. See NOTICE for attribution details.


Siegard Code — Because shipping features shouldn't mean losing control.

Yorumlar (0)

Sonuc bulunamadi