Codeworth — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Codeworth (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.
Aggregate score unchanged between these scans.
The primary manifest — the file an agent reads to learn what this artifact does.
A Claude Code skill that analyzes software repositories and estimates the human time, team composition, and cost required to rebuild them from scratch.
AI-powered repository valuation and rebuild-effort estimation for technical diligence, rewrite scoping, and engineering cost analysis.
Codeworth is a Claude skill/plugin that analyzes a software repository and estimates the human time, team composition, and cost required to rebuild it from scratch.
Unlike naive estimators that rely on lines of code alone, Codeworth evaluates the hidden drivers of engineering effort: architecture, domain complexity, integrations, delivery artifacts, operational maturity, and product overhead.
Most repositories are judged badly.
People look at LOC, frameworks, or a quick skim of the folder tree and pretend they understand replacement cost. They usually don’t.
A small codebase can encode months of specialized security research, ugly integration workarounds, or domain knowledge that took real product and engineering effort to uncover.
Codeworth exists to answer a more useful question:
What would it actually take to rebuild this repo properly?
That makes it useful for:
Codeworth architecture
Codeworth runs two parallel scanners—source analysis and artifact analysis—then uses Claude plus reference guides to produce a rebuild-effort and cost estimate.
Below is a sanitized excerpt from a real Codeworth analysis report. It shows the type of executive-facing estimate produced by the tool without exposing sensitive implementation details from the analyzed repository.
Example analysis: specialized Android security application Estimated rebuild effort: 40–54 weeks Estimated cost range: $350k–$640k USD Confidence: Medium
Although the codebase was modest in size, Codeworth identified several drivers of increased rebuild effort:
Senior Android Engineer
Owns platform architecture and high-risk technical constraints.
Mid-Level Android / Full Stack Engineer
Handles UI flows, integrations, and supporting systems.
Product Manager / Security-Oriented PM
Responsible for threat-model translation, specifications, and release coordination.
| Scenario | Total Hours | Estimated Cost |
|---|---|---|
| Low | ~2,740 hrs | ~$350,000 |
| Mid | ~3,660 hrs | ~$480,000 |
| High | ~4,940 hrs | ~$640,000 |
The codebase appeared relatively small, but the replacement cost was driven by embedded domain knowledge, platform-specific engineering constraints, and product rigor, not raw code volume.
Rebuild cost is driven by hidden complexity, not just code volume.
Codeworth evaluates rebuild difficulty across five dimensions:
Specialized domains like payments, healthcare, cryptography, robotics, distributed systems, trading, tax, ML, or security.
Kubernetes, Terraform, Helm, service mesh, data platforms, deployment pipelines, and operational glue.
Schema depth, ORM model count, migration volume, and underlying state complexity.
External APIs, enterprise systems, billing/auth providers, and custom system interfaces.
SLOs, runbooks, load tests, canary deploys, tracing, compliance artifacts, and production readiness signals.
The result is not just hours. It is a structured estimate with reasoning.
Running Codeworth generates a report containing:
Codeworth combines two layers of analysis.
The scanner inspects:
The artifact scanner detects:
Claude then synthesizes these signals into a structured rebuild estimate.
Workflow:
codeworth/
├── .claude-plugin/
│ └── plugin.json
├── SKILL.md
├── README.md
├── LICENSE
├── evals/
│ └── evals.json
├── references/
│ ├── complexity-guide.md
│ ├── rate-cards.md
│ ├── multipliers.md
│ ├── pm-effort-guide.md
│ ├── log-validation-analysis.md
│ └── rebuild-difficulty.md
└── scripts/
├── analyze_repo.py
└── scan_logs_and_validation.pyCodeworth will be available through the pm-discipline Claude Code marketplace, which will aggregate several product-management-focused skills.
Once available:
/plugin marketplace add teterouge/pm-discipline
/plugin install codeworthClone the repository and load it into Claude Code:
git clone https://github.com/teterouge/codeworth.git
cd codeworthThen add the skill to your Claude environment using your preferred plugin workflow.
Example prompts:
Estimate this repo
Estimate the cost to rebuild this repo from scratch: /path/to/repo
Estimate this GitHub repository for acquisition diligence: github.com/org/repo
Providing context such as team size, region, and purpose will improve results.
This project demonstrates:
It is also useful for founders evaluating replacement cost or technical leverage.
Codeworth analyzes visible repository artifacts and cannot see:
It provides estimates, not formal quotes.
Planned improvements:
MIT
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.