Run bounded iterative refinement loops on PRD, TDD, and tasks-plan artifacts.
npx skills add https://github.com/nodatall/primedirective --skill plan-refine-nodatall
Please help me install this Agent Skill. Skill: plan-refine Source: https://github.com/nodatall/primedirective/tree/main/skills/plan-refine Command: npx skills add https://github.com/nodatall/primedirective --skill plan-refine-nodatall
This skill solves the problem of shipping flawed planning artifacts that lead to implementation drift, missed requirements, unexpected execution blockers, and wasted engineering effort by running a bounded iterative refinement loop with dedicated challenger and reviewer subagents to surface and resolve issues before work begins.
Invoke the plan-refine skill to run a bounded iterative refinement loop over your existing PRD, TDD, and tasks-plan artifacts to catch all blocker and material issues before starting implementation.
High-intent search queries and answers about installing and using this skill.
Pre-execution plan refinement is a structured critique process for PRD, TDD, and tasks-plan artifacts. You need it when you have complete planning artifacts and want to surface hidden assumptions, conflicting requirements, and sequencing issues before implementation begins.
Run a bounded iterative refinement loop with dedicated challenger and reviewer subagents over your PRD and tasks-plan. This pressure-tests planning artifacts for overlooked edge cases and failure paths, automatically stopping when a fresh reviewer round finds no blocker or material issues.
Yes, automated plan critique supports configurable plan-key selection for multiple task plan sets. It automatically preserves research-backed and Pro-backed planning decisions, running final carry-forward checks to ensure no critical context is lost before handoff to execution workflows.
The best way to pressure-test TDD validation artifacts is spawning fresh challenger subagents to identify hidden assumptions and failure paths, paired with reviewer subagents that adjudicate findings and enforce stop rules, running up to 8 structured critique rounds.
Pre-execution review supports configurable round limits up to a maximum of 8 refinement rounds, with a hard stop when no issues are found. It also offers optional refinement log preservation, alongside enforced stop rules and challenge disposition tracking.
Planning artifact review needs dual subagent lanes because dedicated challengers pressure-test for implementation drift traps while dedicated reviewers adjudicate findings and enforce stop rules. This bounded iterative approach resolves issues a single review pass would miss, eliminating execution risk from flawed planning.
rstacruzValidates requirements and scope to identify ambiguities in project plans.
executiveusaEvaluate FX carry trade opportunities using forward curves, volatility surfaces, and interest-rate differentials.
rhixecompanyGuide users through structured co-authoring of technical documents.
v1bh0rAssess architectural concerns in ADRs and architecture documents using structured critique dimensions.
