Works with the AI Agents you already use
Java Legacy Analyzer
Turns a legacy Java codebase into a deterministic inventory, one technical document per class and a dependency-ordered migration plan, without leaking secrets.
$15
Java Legacy Analyzer
Example session with this skill installed
Document the legacy code in src/main/java and give me a migration plan to Java 21 + Spring Boot.
- Read your context and instructions
- Compiled the java legacy analyzer
legacy-analysis/facts.json (inventory, dependency graph, cycles, risk signals), legacy-analysis/classes/<ClassName>.md (one per class: facts, responsibilities, confirmed risks with file:line, migration notes), legacy-analysis/migration-plan.md (risks to remove first, dependency-ordered steps, characterisation tests, S/M/L size classes), plus a summary of what was not analysed and the three highest risks.
Connects securely to your tools. The creator never sees your data.
About this skill
Legacy Java (EJB, Struts 1, raw JDBC, hand-written singletons) is hard to document because nobody remembers how it fits together, and an AI that guesses makes it worse. This skill separates what is measured from what is interpreted:
- A standard-library Python script extracts the structure deterministically: types, fields, methods, dependencies, dependency cycles and legacy risk signals (SQL built by concatenation, swallowed exceptions, manual resource closing, singletons, legacy dates and collections, EJB/JNDI, Struts 1, hard-coded credentials). Same input, same output.
- The agent then reads only what needs judgment and writes one document per class, with facts and analysis kept apart and every claim tied to a file and line.
- The migration plan follows the dependency graph (dependencies first, cycles grouped) and lists the characterisation tests to write before changing anything.
- A redaction script masks passwords, keys, JDBC credentials, e-mails and internal IPs before any excerpt is quoted.
- Built-in honesty rules: read-only on your sources, empty analyses reported as failures, no invented effort estimates or scores.
The detection patterns in the scripts are rules, not secrets: the package contains no credential.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse All Claude Skills.
Regex-based extraction, not a full Java parser: method parameter and return types are not counted as dependencies, and a member declared on the same line as its class's opening brace may be missed. Risk signals are hints that the agent must confirm in the code. Secret redaction has known false negatives (secrets built at runtime, neutral variable names). No effort estimate in days is produced: size classes S/M/L only.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 17 days ago
- Software
Python 3.8+
Permissions
File Scopes
No network access. Scripts use the Python standard library only.