Skip to content

Post-Launch Operations

Internal maintainer playbook for running sopkb day-to-day after the open-source launch. This is not user-facing documentation — it's the checklist maintainers follow to keep the project healthy.

Issue / PR Triage Cadence

Pick a cadence and stick to it.

Recommended cadence: weekly (biweekly is the absolute floor)

At least while the project is small, weekly is recommended — do not go longer than biweekly.

Stale, unattended issues and PRs are the fastest way for a new OSS project to look abandoned. A project with a two-day-old commit and a three-month-old unanswered issue reads as dead to a new contributor, regardless of how active development actually is. Consistency matters more than thoroughness here: a quick weekly pass beats a thorough pass that happens every couple of months.

What "done" looks like for a triage pass: - Every open issue and PR touched since the last pass has been looked at — none left completely untouched. - Each item is either labeled, assigned, or explicitly deferred (e.g. commented "not planned for now, revisiting after vX.Y" and labeled accordingly). - Duplicate or stale items are closed with a short note rather than left to rot. - Anything that needs a maintainer decision is flagged as such so it doesn't silently stall.

Security Reports

SLA: acknowledge within 3 business days

Acknowledge every security report within 3 business days, even if the actual fix takes longer. The acknowledgment is a status update, not a resolution — "we've seen this, we're investigating, we'll follow up by [date]" is sufficient to stop the clock on the SLA.

This cadence pairs with (but does not replace): - The disclosure process described in SECURITY.md (owned separately — reference it, don't duplicate it here). - GitHub Security Advisories for coordinated disclosure and, where applicable, CVE assignment.

If a fix is going to take longer than the initial acknowledgment implied, send a follow-up update rather than letting the reporter wonder whether the report went into a void.

Roadmap Cadence

Revisit and republish the public roadmap every 4-6 weeks. See ROADMAP.md for the current roadmap content (owned by another workstream — this document only sets the cadence for revisiting it, not its content).

The point of the cadence is to prove the roadmap is a living document, not a one-time launch announcement. A roadmap that hasn't moved in three months signals a stalled project even if work is happening elsewhere (commits, releases, triage). Each revisit should at minimum: - Move completed items off (or mark them done). - Reflect any priority changes based on issue/PR volume from triage. - Get republished (announcement, pinned issue update, or discussion post) so the update is visible, not just committed to the repo.

Summary Cadences

Activity Cadence Notes
Issue/PR triage Weekly (biweekly floor) See "done" checklist above
Security report acknowledgment Within 3 business days Pairs with SECURITY.md + GitHub Security Advisories
Roadmap revisit/republish Every 4-6 weeks Roadmap content lives in ROADMAP.md