Execute AI-generated implementation plans with isolated task verification and crash-safe resume.
npx skills add https://github.com/miles990/forge --skill forge-miles990
Please help me install this Agent Skill. Skill: forge Source: https://github.com/miles990/forge/tree/main/skills/forge Command: npx skills add https://github.com/miles990/forge --skill forge-miles990
This Skill provides a robust framework for executing AI-generated plans, ensuring safety, resumability, and collaboration across multiple AI sessions and repositories. It prevents common AI coding failures like lost progress, unsafe code execution, and uncoordinated changes.
Use the forge skill to execute the plan located at docs/plans/my-feature.md.
High-intent search queries and answers about installing and using this skill.
Executing AI-generated implementation plans safely requires isolating tasks in separate Git worktrees, verifying code independently, and merging only passing changes to the main branch. This approach handles multi-repo coordination while ensuring code quality through per-task verification and automatic retries.
Resuming an interrupted AI coding workflow is possible with crash-safe resume capabilities that automatically pick up where a previous run left off. This prevents lost progress by tracking task completion and state across execution sessions.
Collaborative AI development allows multiple AI sessions to work on the same plan concurrently. It coordinates changes across sessions and repositories to prevent uncoordinated modifications and merge conflicts during concurrent execution.
Managing multi-task code execution involves processing AI-generated plans step by step in isolated environments, verifying the code output of each task, and merging only verified changes. This ensures robust safety by retrying failed tasks independently before integration.
Isolating AI code execution tasks uses Git worktrees to create separate working directories for each task. This ensures that code verification and execution happen independently without affecting the main branch until changes pass verification.
AI coding plan execution fails to merge changes safely without isolation because unverified code can directly impact the main branch. Isolating tasks in worktrees and verifying code independently ensures only passing changes are merged, preventing unsafe code execution.
RamblurrDispatch a code-reviewer subagent to verify implementations against requirements.
takemi-ohamaCreate a deploy branch and PR from a feature branch to a target environment.
julianobarbosaReview code for quality, bugs, security issues, and best practices using Codex.
ExpertVagabondChain agent prompts sequentially to execute multi-step workflows.
