Execute hotl-workflow-*.md files in loops until success criteria are met.
npx skills add https://github.com/yimwoo/hotl-plugin --skill loop-execution
Please help me install this Agent Skill. Skill: loop-execution Source: https://github.com/yimwoo/hotl-plugin/tree/main/skills/loop-execution Command: npx skills add https://github.com/yimwoo/hotl-plugin --skill loop-execution
Automates the looped execution of hotl-workflow-*.md files, iterating steps until the defined success criteria are met, and pausing for high-risk gates while auto-approving low-risk gates.
Place a hotl-workflow-<slug>.md in your project root and run the loop-execution skill to start autonomous looping.
High-intent search queries and answers about installing and using this skill.
Looped execution of multi-step workflows is automated by continuously iterating through defined steps until success criteria are satisfied. The runtime reads workflow files, applies frontmatter-driven intent, and pauses execution only when encountering high-risk gates requiring human intervention.
Gate management during human-in-the-loop workflow execution automatically approves low-risk gates to maintain flow. High-risk gates trigger a pause for human review, ensuring safety while allowing the rest of the automated execution to proceed uninterrupted.
Interrupted workflow runs can be resumed using durable state persistence. The runtime captures state snapshots during execution, allowing the looped process to recover and continue from the exact point of interruption without losing prior progress.
An execution report for auditing automated workflow loops is produced automatically upon completion or interruption. The runtime generates durable execution reports and state snapshots to enable debugging, auditing, and resuming workflows.
To start automating hotl-workflow files, place a hotl-workflow-<slug>.md file in your project root. Running the automation skill reads this file to begin autonomous looping through the defined workflow steps and gates.
yimwooGenerate hotl-workflow-<slug>.md implementation plans with tasks, loops, and gates.
andershi666Records and inspects agent execution snapshots, including messages, context, and tool calls, via CLI.
DhumitechBuild graph-based AI agent workflows with conditional routing and persistent state.
yimwooExecute implementation plans with human checkpoints between task batches.
