What problem does it solve?
Producing reliable TechSpecs is slower when you need cross-system reference grounding, because competitor repos must be studied thoroughly and converted into a consistent, actionable format.
Core Features & Use Cases
- Dispatches scoped-write competitor subagents in parallel: runs the dedicated
cy-researcher subagent once per competitor, each writing exactly one analysis file to the task’s analysis directory.
- Enforces a fixed analysis schema with validation: uses a canonical template to ensure every analysis includes the required seven sections and that evidence cites real file paths under the competitor’s
.resources/<name>/.
- Produces reusable architecture inputs: generates findings you can link into
_techspec.md or task artifacts while keeping the analysis files as the authoritative source.
Use this when you need cross-system reference grounding for a TechSpec, refactor, or design decision that depends on patterns from multiple existing agent runtimes and harnesses.
Quick Start
Ask the AI to run the competitor research for your task by providing the task slug and a comma-separated list of competitor names, for example: Dispatch cy-research-competitors for task "my-feature" and competitors "hermes,openclaw,openfang".
