Summary
There is no skill in the registry for managing LibreFang workflows (create, list, run, inspect runs, delete). Agents that need workflow capabilities currently have librefang-workflow in their skills list, but the skill doesn't exist — it resolves to nothing at runtime.
Motivation
LibreFang has a full workflow engine (crates/librefang-kernel/src/workflow.rs) with API endpoints for CRUD operations, step execution, and run history. Agents should be able to discover and use these capabilities through a skill that provides:
- Guidance on
workflow_run, workflow_list, workflow_create tools
- Best practices for multi-step workflows (error handling, human-in-the-loop steps)
- Examples of chaining agents, conditions, and parallel execution
- Context about the workflow editor in the dashboard
Current state
- Workflow tools exist as builtins (
workflow_run, workflow_list, etc.)
- Dashboard has a workflow editor with visual step builder
- No skill provides the prompt context to help agents use workflows effectively
librefang-workflow is referenced in agent configs but doesn't resolve
Suggested scope
A librefang-workflow skill with a SKILL.md that describes available workflow tools, their parameters, and usage patterns — similar to how git-expert provides context for git operations.
Summary
There is no skill in the registry for managing LibreFang workflows (create, list, run, inspect runs, delete). Agents that need workflow capabilities currently have
librefang-workflowin theirskillslist, but the skill doesn't exist — it resolves to nothing at runtime.Motivation
LibreFang has a full workflow engine (
crates/librefang-kernel/src/workflow.rs) with API endpoints for CRUD operations, step execution, and run history. Agents should be able to discover and use these capabilities through a skill that provides:workflow_run,workflow_list,workflow_createtoolsCurrent state
workflow_run,workflow_list, etc.)librefang-workflowis referenced in agent configs but doesn't resolveSuggested scope
A
librefang-workflowskill with a SKILL.md that describes available workflow tools, their parameters, and usage patterns — similar to howgit-expertprovides context for git operations.