_bmad-shared/skills/gd-shared-manage-epic/SKILL.md
Autonomously drives all stories in one or more epics using single-purpose subagents. Use when the user says "manage epic [N] autonomously", "run epics [N,M] in autopilot", or "manage all epics".
npx skillsauth add grasdouble/bmad-manager gd-shared-manage-epicInstall 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.
This skill autonomously manages the full lifecycle of stories across one or more epics — create, dev, review, and retrospective — by orchestrating dedicated single-purpose subagents. Act as the user's proxy: give each subagent exactly one action, collect its result, then advance. No subagent ever handles more than one action type.
Args:
epic 1 or just 1epic 1 2 3 or 1,2,3all — the skill reads sprint-status.yaml and processes every non-done epic in orderLoad config from {project-root}/_bmad/config.yaml and {project-root}/_bmad/config.user.yaml (bmm section). Resolve:
communication_language, implementation_artifacts, planning_artifactssprint_status path = {implementation_artifacts}/sprint-status.yamlParse the epic target from the invocation argument:
sprint-status.yaml, collect all epic keys (epic-N) whose status is not done, sort ascending — that is the queuePresent the epic queue to the user and confirm before starting. Then load references/orchestration.md to begin processing each epic in sequence.
documentation
Maintains up-to-date documentation for a standard (non-monorepo) project. Use when the user says 'update project docs', 'update documentation', 'mise à jour la doc' or 'mettre à jour la documentation'.
documentation
Maintains up-to-date documentation for a monorepo project using parallel subagents. Use when the user says 'update project docs', 'update documentation', 'mise à jour la doc' or 'mettre à jour la documentation'.
development
Lists all monorepo packages grouped by category with versions. Use when the user wants to see packages, check versions, or asks for a package inventory.
testing
Push the LLM to reconsider, refine, and improve its recent output. Use when user asks for deeper critique or mentions a known deeper critique method, e.g. socratic, first principles, pre-mortem, red team.