From autoforge
Operate AutoForge as a durable long-running agent runtime with queueing, watch mode, async messages, checkpoint replies, and resumable execution.
How this skill is triggered — by the user, by Claude, or both
Slash command
/autoforge:long-run-runtimeThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill when the user needs a workflow that should continue beyond the current Claude session.
Use this skill when the user needs a workflow that should continue beyond the current Claude session.
AutoForge supports:
Canonical runtime flow:
autoforgeai daemon startautoforgeai queue "<objective>" --wait --tailautoforgeai projectsautoforgeai watch <project_id> --tailautoforgeai msg <project_id> "<updated guidance>"autoforgeai unpause <project_id>Use this mode whenever the user cares about:
npx claudepluginhub alyciabhz/autoforge --plugin autoforgeExecutes long-running tasks autonomously across Claude Code sessions using headless bash loops or in-session hooks. Supports structured decomposition for projects and Ralph-style iteration for TDD, fixes, refactoring.
Framework for long-running, multi-session autonomous agent tasks with progress checkpointing, failure recovery, and task dependency management. Maintains append-only progress logs and checkpoint files for recovery across context window resets.
Orchestrates long-running tasks in Claude Code by routing through /goal, ~auto, and ~qa for continuous execution, cross-session recovery, and quality closure.