framework-generator — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited framework-generator (Plugin) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
A tool for creating structured, repeatable frameworks with executable skills through staged refinement. Works with Claude Code.
git clone https://github.com/athrael-soju/Framework-Generator.git
cd framework-generator
claudeClaude Code will automatically read the project's CLAUDE.md for context and instructions.
Once in a Claude Code session, ask Claude to execute stages:
| Request | Description |
|---|---|
| "Run Frame stage for {name}" | Define framework purpose and boundaries |
| "Run Organize stage for {name}" | Map stages and flow |
| "Run Refine stage for {name}" | Specify each stage in detail |
| "Run Generate stage for {name}" | Produce framework artifacts |
| "Run Evaluate stage for {name}" | Validate and iterate |
See the commands/ directory for detailed stage instructions.
┌─────────┐ ┌──────────┐ ┌────────┐ ┌──────────┐ ┌──────────┐
│ Frame │───▶│ Organize │───▶│ Refine │───▶│ Generate │───▶│ Evaluate │
└─────────┘ └──────────┘ └────────┘ └──────────┘ └──────────┘
▲ │
└────────────────── feedback loops ──────────────────────────┘Run Frame stage for my-workflowYou'll be asked about:
Output: Framework charter in output/{date}/my-workflow/1-frame/charter.md
Run Organize stage for my-workflowDefine:
Output: Stage map in output/{date}/my-workflow/2-organize/stage-map.md
Run Refine stage for my-workflowFor each stage, specify:
Output: Stage specs in output/{date}/my-workflow/3-refine/{stage}-spec.md
Run Generate stage for my-workflowProduces:
Output: Complete framework in output/{date}/my-workflow/4-generate/my-workflow/
Run Evaluate stage for my-workflowValidates:
Output: Validation report in output/{date}/my-workflow/5-evaluate/validation.md
output/{date}/{framework-name}/
├── run.md # Progress tracking
├── 1-frame/
│ └── charter.md # Framework charter
├── 2-organize/
│ └── stage-map.md # Stage map with flow diagram
├── 3-refine/
│ └── {stage}-spec.md # Stage specifications
├── 4-generate/
│ └── {framework-name}/
│ ├── .claude/
│ │ └── settings.local.json # Claude Code permissions
│ ├── .claude-plugin/
│ │ └── plugin.json # Plugin manifest
│ ├── commands/
│ │ └── {stage}.md # Stage command files
│ ├── README.md # Framework overview
│ └── CLAUDE.md # AI instructions
└── 5-evaluate/
└── validation.md # Validation reportEvery stage requires approval before proceeding. This ensures:
Each stage has explicit checklists:
Built-in iteration paths:
Generated frameworks include everything needed to run independently:
.claude/settings.local.json for permission configurationGenerated frameworks include .claude/settings.local.json with bypassPermissions enabled. This allows Claude Code to execute commands without prompting for approval at each step.
Risks:
Alternatives:
.claude/settings.local.json to restore default permission prompts"defaultMode": "default" for selective approval| Type | Pattern | Example |
|---|---|---|
| Foundation | Single assessment, run once | Initial analysis |
| Pipeline | Sequential stages, ongoing | Feature development |
| Cyclical | Repeating on cadence | Weekly reviews |
| Hub | Central stage, others connect | Decision framework |
MIT
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.