tachun/shape-up-analysis-skills
Overview
This skill analyzes a Shape Up pitch alongside the codebase to produce a concrete kickoff analysis with identified risks, rabbit holes, and a bet confidence rating. It delivers a practical, prioritized set of spikes, open questions, and mitigations to enable a focused cycle that fits the appetite.
How this skill works
I read the Shape Up pitch to extract problem, appetite, solution, and explicit no-gos, then map those items to the actual code paths that will be affected. I assess code health, surface hidden complexity, categorize risks (known risks, rabbit holes, unknowns), and synthesize a kickoff document with a numeric/star confidence rating and actionable next steps.
When to use it
- When starting a new Shape Up cycle and you need to validate scope against codebase reality
- When someone asks to "analyze this pitch" or "review Shape Up pitch"
- Before betting: to surface blockers, unknowns, and rabbit holes
- When you need a kickoff document with spikes, decisions, and confidence
- To translate product assumptions into concrete technical questions
Best practices
- Provide the full pitch document and grant read access to the code areas touched by the pitch
- Keep the appetite explicit so risks are evaluated against a fixed timebox
- Accept trade-offs up front and mark no-gos clearly to reduce scope creep
- Run short timeboxed spikes for high-likelihood/high-impact unknowns before committing
- Prioritize open questions by if they block progress (Can Proceed? flag)
Example use cases
- Evaluate a user-facing feature pitch that touches payments and auth to find hidden integration work
- Review an internal tooling pitch to discover legacy code hotspots that could explode scope
- Prepare a kickoff summary for leadership showing confidence, blockers, and required decisions
- Identify spikes required to estimate effort within a two-week appetite
- Assess whether a pitch meets betting criteria and what would raise/lower confidence
FAQ
A complete Shape Up pitch (problem, appetite, solution, rabbit holes, no-gos) and read access to the repository paths the pitch affects.
How is bet confidence determined?
Confidence is synthesized from matched assumptions vs code reality, risk likelihood × impact, presence of tests/docs, and whether key unknowns are spike-able within the appetite.