--- type: "post" id: "claude-plugins-official-plugins-code-modernization-readme-md-000530" slug: "claude-plugins-official-plugins-code-modernization-readme-md-000530" title: "claude-plugins-official/plugins/code-modernization/README.md at main | anthropics/claude-plugins-official" description: "Thesis: Legacy-code modernization becomes more operationally credible when transformation is paired with a separate, inspectable behavioral verification step." retrieval_nugget: "Thesis: Legacy-code modernization becomes more operationally credible when transformation is paired with a separate, inspectable behavioral verification step." published_at: "2026-10-14" updated_at: "2026-10-14" record_date: "2026-10-14" date_kind: "published_at" topics: [] source_urls: ["https://github.com/anthropics/claude-plugins-official/blob/main/plugins/code-modernization/README.md"] schema_version: "newruntime-agent-readable-v0.2" stable_id: "post:claude-plugins-official-plugins-code-modernization-readme-md-000530" status: "published" visuals: [] editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:ffc932d88d4ac02ff7f21363d42e7b2d60279d1c8f04180641a7e14ed1c710f1","reviewed_at":"2026-10-14T08:02:53.104Z","source_evidence_count":1,"verified_claim_count":3,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":3,"implication_count":2,"watch_condition_count":1,"related_record_count":0} analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Legacy-code modernization becomes more operationally credible when transformation is paired with a separate, inspectable behavioral verification step.","observed_facts":[{"text":"The README lists a workflow from assessment and mapping through reviewed business rules, plan approval, modernization, and verification.","source_urls":["https://github.com/anthropics/claude-plugins-official/blob/main/plugins/code-modernization/README.md"]},{"text":"The README says verification reruns tests from a clean build and compares old and new outputs on the same inputs where the legacy code can run.","source_urls":["https://github.com/anthropics/claude-plugins-official/blob/main/plugins/code-modernization/README.md"]},{"text":"The README says the legacy source is not edited and identifies separate analysis and modernized directories.","source_urls":["https://github.com/anthropics/claude-plugins-official/blob/main/plugins/code-modernization/README.md"]}],"mechanism":"The plugin makes a post-build verification step rerun tests and compare old and new implementations on shared inputs, with listed masks and recorded accepted differences.","why_now":"The plugin makes verification a named stage in an AI-assisted modernization workflow rather than relying on the model’s assertion that a rewrite is correct.","implications":["Teams evaluating AI-assisted migration can assess the verification artifacts and human checkpoints, not just the generated code.","The documented process still depends on buildability, representative inputs, and human decisions; the README does not establish success rates."],"evidence_boundary":"The public README describes product behavior and intended process, but supplies no independent benchmark or outcome data, and explicitly leaves key decisions to people.","watch_conditions":["Watch for published examples showing whether the comparison and test evidence catches meaningful regressions on real migrations."],"related_records":[],"new_branch_reason":"No prior New Runtime record was supplied; this is a distinct plugin release centered on legacy modernization and behavioral verification."} editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]} routes: {"html":"https://newruntime.com/posts/claude-plugins-official-plugins-code-modernization-readme-md-000530/","markdown":"https://newruntime.com/posts/claude-plugins-official-plugins-code-modernization-readme-md-000530.md","json":"https://newruntime.com/posts/claude-plugins-official-plugins-code-modernization-readme-md-000530.json"} --- # claude-plugins-official/plugins/code-modernization/README.md at main | anthropics/claude-plugins-official ## Retrieval answer Thesis: Legacy-code modernization becomes more operationally credible when transformation is paired with a separate, inspectable behavioral verification step. Thesis: Legacy-code modernization becomes more operationally credible when transformation is paired with a separate, inspectable behavioral verification step. Anthropic’s Code Modernization plugin for Claude Code lays out that sequence in its official README: assess and map a system, extract rules, obtain human review and plan approval, then modernize and verify. The README supports a workflow claim, not a claim that migrations are proven safe in general. Mechanism: The plugin makes a post-build verification step rerun tests and compare old and new implementations on shared inputs, with listed masks and recorded accepted differences. It also keeps the legacy source untouched and directs work to separate analysis and modernized directories. These checks make the verification process inspectable, but do not establish measured success rates. Why now: The plugin makes verification a named stage in an AI-assisted modernization workflow rather than relying on the model’s assertion that a rewrite is correct. Evidence boundary: The public README describes product behavior and intended process, but supplies no independent benchmark or outcome data, and explicitly leaves key decisions to people. First implication: Teams evaluating AI-assisted migration can assess the verification artifacts and human checkpoints, not just the generated code. First watch condition: Watch for published examples showing whether the comparison and test evidence catches meaningful regressions on real migrations. Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json ## Continue in the graph Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`. - collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis. - collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph. - whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)