skills/repo-modernizer/SKILL.md
Repo/monorepo modernization: dependency upgrades, security fixes, deprecation cleanup, framework migrations, dependency-native refactors, and verified hard-cut simplification.
npx skillsauth add bjornmelin/dev-skills repo-modernizerInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Use this skill when you need to audit, upgrade, refactor, and simplify a repo or monorepo end-to-end around dependencies, security findings, deprecations, and framework migrations.
Use the base prompt below by default. If the repo shape or workflow calls for a specialized operating mode, load the matching reference first:
references/00-variant-index.mdreferences/framework-lane-selection.mdreferences/dependency-decision-matrix.mdreferences/verification-matrix.mdreferences/report-template.mdreferences/bun-first.mdreferences/pnpm-turbo-monorepo.mdreferences/plan-first.mdreferences/gh-deps-workflow.mdreferences/gh-deps-report-spec.mdSelection guidance:
references/bun-first.mdreferences/pnpm-turbo-monorepo.mdreferences/plan-first.mdscripts/ and
load references/gh-deps-workflow.mdreferences/framework-lane-selection.mdreferences/dependency-decision-matrix.mdreferences/verification-matrix.mdreferences/report-template.mdreferences/00-variant-index.md firstUse these throughout the task:
$bun-dev$hard-cut$clean-code$reducing-entropy$githubcontext7web.runUse $opensrc conditionally as an escalation path for package upgrades
that need source-level proof.
Framework-specific routing is defined in the execution prompt below and refined
further in references/framework-lane-selection.md.
Resolve skill_dir as the directory containing this skill before running the
bundled dependency-intel scripts.
python3 "$skill_dir/scripts/gh_deps_intel.py" full --repo . --out reports --mode safepython3 "$skill_dir/scripts/gh_deps_intel.py" package --repo . --out reports --mode safe --dependency <name>python3 "$skill_dir/scripts/gh_deps_intel.py" scan --repo . --out reportspython3 "$skill_dir/scripts/gh_deps_intel.py" enrich --repo . --out reports --mode safepython3 "$skill_dir/scripts/gh_deps_intel.py" analyze --repo . --out reports --mode safepython3 "$skill_dir/scripts/gh_release_diff.py" owner/repo --current <version> --target <version>python3 "$skill_dir/scripts/gh_compare_notes.py" owner/repo <old-ref> <new-ref>python3 "$skill_dir/scripts/gh_rate_limit_diag.py"Default to --mode safe. Use --mode fast only when the user accepts bounded
parallel API calls. Produce both Markdown and JSON reports.
Load these references when using this lane:
references/gh-deps-workflow.mdreferences/gh-deps-command-mapping.mdreferences/gh-deps-github-api-endpoints.mdreferences/gh-deps-compatibility-policy.mdreferences/gh-deps-report-spec.mdreferences/gh-deps-troubleshooting.mdFor full execution mode, load references/execution-prompt.md. It contains the complete modernization prompt, research requirements, dependency audit workflow, framework lanes, upgrade strategy, verification expectations, and final report contract.
Keep this entrypoint focused on routing and reference selection. Do not duplicate framework-specific playbooks here; use the variant references above.
development
Pre-PR multi-model review, parallel opus and codex exec adversarial lanes, then adversarial verification of merged findings. Read-only. Use before shipping nontrivial diffs.
tools
Independent gpt-5.6 diff review via the Codex CLI, normal or steerable adversarial with JSON findings. Use before shipping nontrivial changes.
development
Delegate implementation, investigation, or bulk work to gpt-5.6 codex via pinned codex exec. Use for clear-spec builds, migrations, debugging, or any task MODELS.md routes to codex.
development
Adversarial pre-mortem: imagine the plan failed, work backwards to surface risky assumptions + irreversible bets, then harden them. Proactively offer it (after the current request; confirm first) before a hard-to-reverse or one-way-door call (API, schema, framework, a hire), an all-upside plan, or unvalidated assumptions. Also on request.