nixos-config

Load this skill when drafting or editing a Jira Change Request (CR) ticket in the CH project (Change Release) at thankyoupayroll.atlassian.net, or any change ticket whose primary reader is an executive approver rather than the implementing engineer. The skill covers audience framing, description structure, field-by-field guidance, and the CH project's known quirks (ADF formatting, character limits).

Spawn isolated agent sessions in their own git worktrees using the prism tool. Use when the user asks to spawn an agent, delegate work to another session, run something in parallel, or work on a PR or different repo.

Load this skill when the user signals that the current branch is a long-lived work-in-progress (WIP) branch where external review is deferred — phrases like "WIP", "work in progress", "long-lived branch", "feature branch", "defer review", "no reviews yet", "treat as WIP", "don't run reviews", "skip review for now", or similar. Once loaded, the worker stops kicking off `prism review` after each push and ignores the git-push reminder's review instruction until the user explicitly signals readiness.

Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.

Load this skill when reading from or writing to Jira Cloud or Confluence Cloud — searching issues, fetching a specific Jira issue or Confluence page, listing or performing workflow transitions, creating/updating issues, commenting on issues, publishing or updating Confluence pages, or understanding the atlassian MCP tools.

Load this skill when drafting or reviewing acceptance criteria for an issue or ticket — before creating a work item, preparing a worker spawn prompt, or critiquing existing ACs before accepting work.

Use AWS CLI correctly. Load this skill when doing any AWS work — querying resources, managing profiles, assuming roles, or asking the user to run AWS commands.

Load when the user asks for a retrospective, postmortem, or after-action analysis of a specific agent session (and its review cycles) — phrases like "do a retro of that session", "postmortem this", "how did `<session>` go", "what went wrong in that session", or "analyse the review cycles for that PR". Targeted analysis of one session group (worker + its `~review-N-<agent>` subsessions); not for broad-window sweeps across many sessions.