{"slug":"claude-code-dependency-upgrade","title":"Upgrade one dependency and read the changelog first","summary":"Claude Code summarizes breaking changes for a single dependency, applies the upgrade, and runs the existing tests. Major bumps stay behind a question.","agent":"Claude Code","agentSlug":"claude-code","category":"Coding","categorySlug":"coding","roles":["Software engineer"],"roleSlugs":["software-engineer"],"industries":["Software"],"industrySlugs":["software"],"jobs":["shipping-code"],"topic":"Dev & Coding","topicSlug":"dev-and-coding","score":22,"difficulty":"hard","estCost":"Claude Code on a paid Claude plan. CI minutes are your own.","costTier":"low","timeSaved":"The changelog reading and the mechanical codemod, not the production watch afterward.","prompt":"Upgrade only the dependency I name. Do not bump unrelated packages to \"make the lockfile green\" unless the package manager forces it. If it does, list those bumps separately.\n\nBefore editing:\n- Read the changelog or release notes between the installed version and the target.\n- List breaking changes that could touch this repo, with a file path if you can find a call site.\n- If the target is a new major version, stop and ask me to confirm.\n\nThen upgrade, run the test and typecheck scripts in package.json or the equivalent, and paste failures with the first useful stack line. Do not silence types with a blanket ignore.","steps":[{"title":"Pick one package and a target version","detail":"Write both in the prompt. 'Upgrade everything' is how lockfiles become unreviewable."},{"title":"Ask for the breaking-change list first","detail":"You can split this into two turns if the model rushes. The list is the point of the session."},{"title":"Let it run the repo's own checks","detail":"Claude Code can execute the test script. Read the output rather than trusting a summary that says all green."},{"title":"Ship the upgrade in its own pull request","detail":"Keep refactors out of the same diff so a rollback is possible."}],"sourceUrl":"https://code.claude.com/docs/en/quickstart","sourceNote":"Official documentation URL, checked when this seed was written. The prompt and steps are an original AI Playbook Gems playbook, not a quotation of a customer story, and have not been re-run.","date":"2026-10-12","verified":false,"status":"unverified","sponsorUrl":"","metrics":[{"label":"Hands-on time","value":"Longer than a typo fix; plan on a review session"},{"label":"You still approve","value":"Major-version confirmation and the merge"},{"label":"Cost bucket","value":"Low"}],"faq":[{"question":"Has this Claude Code playbook been re-run for the catalog?","answer":"No. Status is unverified. Paste the prompt, keep the version that worked, and do not treat the time or cost lines as measurements."},{"question":"What if the changelog is missing?","answer":"The prompt tells the agent to stop on a major bump and to list call sites it can find. You decide whether the gap is acceptable."}],"url":"https://aiplaybookgems.com/usecases/claude-code-dependency-upgrade","jsonUrl":"https://aiplaybookgems.com/api/usecases/claude-code-dependency-upgrade"}