Use cases
Jobs where this local tool earns its keep. Each names the role, the task before the deadline, the steps, and what done looks like.
1. CI engineer rewriting internal trigger scripts
Who: A CI engineer who owns internal scripts and glue that still POST CircleCI /api/v1 or /api/v1.1 project build and retry paths.
Task: Finish rewrites before 2027-02-01 so automations do not hit HTTP 410 Gone when the ten legacy trigger and rerun endpoints are removed.
Steps: Grep for the changelog endpoint list; paste matches into the home rewriter; pick cloud or server; apply the map to pipeline/run or v3 runs; switch retries to workflow rerun with SSH flags; smoke a non-prod pipeline; attach Markdown plus Changelog links.
Result: No internal script posts legacy trigger or rerun paths. Canary pipelines succeed on v2 or v3, and the ticket holds the exported rewrite pack.
2. Platform engineer on CircleCI server
Who: A platform engineer operating CircleCI server where API v3 is unavailable and only API v2 is the supported migration target.
Task: Migrate server-side triggers to API v2 pipeline/run and workflow rerun without accidentally depending on cloud-only v3 endpoints.
Steps: Select platform=server on the home tool; rewrite username/repo addressing to provider/org/project; map SSH via enable_ssh; document that other unrelated v1 endpoints remain; export the pack for operators.
Result: Server automations survive the 410 cut without a v3 dependency. Operators have a written map of every rewritten caller.
3. On-call responding to 410 after the removal date
Who: An on-call engineer who sees HTTP 410 Gone from an old integration after February 1 2027.
Task: Identify which legacy path fired and hand a precise rewrite pack to the owning team instead of a vague “upgrade CircleCI” comment.
Steps: Paste the failing path into the rewriter; generate the cloud/server map; open a ticket with Changelog links; temporarily disable the caller if needed; schedule the durable rewrite with the owner.
Result: Incident notes include the exact replacement endpoint and SSH/rerun flag guidance so the next deploy does not reintroduce build_num retry.
How-to and FAQ.