What problem does it solve?
Editing Warp software factory definitions (factory.yaml, Agent, Automation, Scorer, Runner, and Webhook files) is error-prone because the format is owned by warp-server and changes over time. This Skill guides correct authoring of those files and validates the tree against the live server parser, avoiding stale local schema copies that report wrong diagnostics.
Core Features & Use Cases
- Schema-guided authoring: Fetches the exact JSON schema for the tree's declared schemaVersion from the warp-server schema endpoints before editing any resource file.
- Server-backed validation: Runs the bundled Python validator, which submits the tree's resource files to the warp-server validation endpoint and reports real parser diagnostics with FF_* codes, paths, and line numbers.
- Honest validation outcomes: Distinguishes three exit codes (valid, diagnostics found, not validated) and forbids claiming a tree was checked when the server was unreachable.
- Use Case: You add a new Automation with a GitHub pull_request trigger to your factory tree. The Skill reads factory.yaml, fetches the v1alpha1 schema, writes the automation.md with correct trigger filter keys, then runs the validator and fixes every reported diagnostic before you open the pull request.
Quick Start
Ask the agent to create or edit a factory resource file in the repository rooted at factory.yaml and validate the tree with the bundled validation script before opening a pull request.