Create and execute validated YAML workflow recipes for hatch3r agents.
npx skills add https://github.com/hatch3r/hatch3r --skill hatch3r-recipe
Please help me install this Agent Skill. Skill: hatch3r-recipe Source: https://github.com/hatch3r/hatch3r/tree/main/skills/hatch3r-recipe Command: npx skills add https://github.com/hatch3r/hatch3r --skill hatch3r-recipe
Recipes standardize repeatable hatch3r workflows so teams can run complex multi-step agent plans safely and consistently instead of re-deriving the same sequence each time.
depends_on and parallel_with relationships to build an execution graph and run safe parallel substeps when appropriate..hatch3r/recipes/ (project) or ~/.hatch3r/recipes/ (user-level) depending on reuse needs.Create a recipe YAML in .hatch3r/recipes/ and run it in dry-run mode to confirm the DAG, variables, and referenced commands/skills before executing for real.
High-intent search queries and answers about installing and using this skill.
You create repeatable agent workflows by defining a YAML recipe with variables, step DAGs, and checkpoints. This standardizes multi-step agent plans so teams can execute complex sequences safely and consistently without re-deriving the process each time.
The best way to validate a dependency graph is using dry-run mode. Dry-run execution validates prerequisites, detects cycles, catches missing references, and confirms the DAG and variables before you execute the workflow for real.
Yes, you can resume workflow execution after failures. The recipe orchestration supports a recovery workflow with checkpoint guardrails, allowing you to resume safely from the last valid state instead of restarting the entire sequence.
Dependency-aware execution graphs model depends_on and parallel_with relationships to build an execution sequence. This allows the orchestrator to run safe parallel substeps when appropriate based on the constructed dependency graph.
You should store workflow recipes at the project level in .hatch3r/recipes/ for project-specific automation, or at the user level in ~/.hatch3r/recipes/ for global reuse across multiple projects depending on your needs.
yu13130122297Identify uncertainties in research framing, datasets, metrics, and baselines.
hjemmesidekongenOrchestrate multi-stage development pipelines with YAML configuration and state tracking.
klu-devDecompose RFC-specified features into work units for DAG-based multi-agent execution.
z1439527767Detect and recover failed AI agent tool executions with retries and checkpoints.