Structures a time-bound risk spike to evaluate load-bearing unknowns and document decisions.
npx skills add https://github.com/leifericf/agentic-sdk --skill run-spike
Please help me install this Agent Skill. Skill: run-spike Source: https://github.com/leifericf/agentic-sdk/tree/main/skills/run-spike Command: npx skills add https://github.com/leifericf/agentic-sdk --skill run-spike
This Skill helps developers address and mitigate risks associated with unknowns in software development, by structuring a risk spike to evaluate and decide on the next steps.
Run a risk spike for the 'native edge' unknown by executing the 'run-spike' command.
High-intent search queries and answers about installing and using this skill.
A risk spike is a time-bound investigation used to evaluate load-bearing unknowns in software development. Developers build a throwaway harness to test a specific seam, allowing them to decide on go, no-go, or mitigation based on performance and stability.
To evaluate a load-bearing unknown, structure a risk spike with a predefined time-box. Build a throwaway harness to exercise the unknown seam, document failure modes and limits, then record decisions and constraints for future reference.
Run a time-boxed spike when facing a complex native edge or real-time budget constraint. Use this approach to exercise the seam with a throwaway harness and determine next steps based on actual performance and stability rather than speculation.
Document a risk spike outcome by recording the go, no-go, or mitigation decision based on performance results. Include failure modes, identified limits, workarounds, and constraints to provide future reference and rationale for the decision.
The best way to mitigate unknown risks in a real-time budget scenario is executing a structured risk spike. Build a throwaway harness to exercise the seam within a time-box, evaluating performance and stability to determine go, no-go, or mitigation.
Yes, a risk spike requires a throwaway harness to isolate and exercise the load-bearing unknown. This time-bound execution focuses purely on evaluating performance and stability for decision-making, preventing code entanglement with production systems.
lucasgaldinosCreate structured spike documents with frontmatter in docs/spikes.
mietheGenerate PRDs and phased implementation plans with subagent assignments.
ScotterMonkCreate, read, and write timestamped log entries for project plans.
zangjeicyRun disposable technical experiments to validate feasibility and compare implementation approaches.
