SssoGin
UserMulti-Agent Long-Task Scheduling and Growth System for AI coding agents, with file-based state, verification, handoff, and reusable learning.
Categories
Indexed Skills (6)
grill-me-preflight
Use before non-trivial task or project starts to expose hidden assumptions, goal boundaries, key tradeoffs, and acceptance criteria.
malts-project-init
Initialize or refresh MALTS-aware project instructions for a workspace. Use when the user asks for an /init-like workflow, project initialization, project-level AGENTS.md, PROJECT_CONTROL.md, 项目控制.md, WORK_TASK_REPORT.md, 工作任务报告.md, Claude Code compatibility through CLAUDE.md, or setup of MALTS operating rules for a new or existing project. This is a shared MALTS installed skill for supported Agent adapters, not a Claude Code slash command.
multi-agent-long-task-scheduling
Use when the user explicitly enables multi-agent long-task scheduling, or when a complex long task needs recoverable state, task contracts, sub-agent dispatch, verification, and delivery control.
project-retrospective-growth
Use after project completion, phase delivery, rework, user correction, verification failure, or explicit retrospective requests to distill reusable experience without polluting long-term memory.
single-agent-lightweight-growth
Use by default during normal single-agent work to keep growth continuous but cheap, without enabling full multi-agent scheduling.
session-handoff
Use when the user asks for a project handoff, next-Agent summary, continuation notes, PROJECT_HANDOFF, or similar recovery context.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.