Docs-Anchored Plan Stress Test
Stress-test plans against project terminology and ADRs, then reconcile accepted decisions with precise document edits.
Before & after
Same prompt, same agent — the only difference is the skill.
Splitting a monolith is a significant decision with tradeoffs. There are good reasons to do it, like independent scaling and team autonomy, but it introduces complexity in distributed systems, operational overhead, and testing challenges. It really depends on your team size, current pain points, and long-term vision. Do you have specific constraints driving this question? What parts feel like they need to be separate?
DECISION ANALYSIS vs. EXISTING DOCS CONFLICT WITH ADR-003 (Monolith-First Policy) Doc states: Stay monolith until services exceed 200k lines or 8+ teams Current state: 156k lines, 4 teams Conclusion: Violates established threshold policy Action Required: Update ADR-003 or defer split ALIGNMENT CHECK vs. CONTEXT.md Documented values: "Operational simplicity over microservices cleverness" Current proposal: Splits payment + user services Risk: Creates 3 database schemas, 2 new async queues Missing from proposal: Operational runbook for split deployments CRITICAL GAPS 1. ADR-003 suggests deferral; new pain points not yet in docs 2. CONTEXT.md emphasizes observability first; no tracing plan shown 3. Team structure doc (outdated) assumes monolith ownership model UPDATES TO WRITE • ADR-004: Justify split against original threshold criteria • CONTEXT.md: Add "Distributed systems" as new value if proceeding • Team Structure: Define on-call model for split services RECOMMENDATION: Refactor payment service to separate package first; defer full split until ADR reviewed
About this skill
Your migration plan looks reasonable, but it quietly changes what “tenant” means and reverses an architecture decision nobody mentioned. This skill catches those conflicts, asks the question that resolves each one, and turns accepted answers into precise plan and documentation edits.
What it does
- Checks plan terminology against CONTEXT.md, including overloaded terms and changes to identity, ownership, or lifecycle.
- Traces proposed implementation choices to relevant architecture decision records, including their status, scope, and exceptions.
- Distinguishes actual contradictions from undocumented choices, stale records, and decisions that no longer apply.
- Challenges the highest-impact conflict first with a cited passage, concrete consequence, and focused question.
- Updates authorized documentation as decisions settle, preserving ADR history and recording unresolved gaps.
What you get
A focused review session with evidence-backed findings, reconciled plan language, and precise documentation changes. The closing record identifies what was resolved, what remains open, and which edits were actually applied and verified.
Best for
Engineers and technical leads reviewing migrations, feature plans, domain-model changes, or architecture proposals in projects that maintain CONTEXT.md and ADRs.
SKILL.md preview
---
name: docs-anchored-plan-stress-test
description: Stress-tests a plan against CONTEXT.md terminology and relevant architecture decision records, citing conflicts and reconciling accepted decisions through precise documentation edits. Use when the user says "grill this plan against our docs", "does this contradict our ADRs?", "check this migration against CONTEXT.md", "our plan uses tenant differently", "the architecture docs and proposal disagree", "we keep reopening decisions we already made", or "update the docs as we settle this".
---
# Docs-Anchored Plan Stress Test
… (sign up to view the full skill)More development skills
View all Development skills →Performance Bottleneck Profiling: Node, Python, Go
Classifies a slowdown, gives exact profiler commands, reads the output, and proves the fix with same-conditions p95/p99 numbers.
Feature Scaffolding with Working Tests
Turn a short feature spec into code, runnable tests, and integration wiring that match your existing project.
Feature Flag Rollouts and Kill Switches
Design staged flag rollouts, verify rollback paths, and audit flag debt with explicit defaults, gates, and retirement plans.
Microcontroller Firmware and RTOS Engineering
Build and debug MCU firmware with bounded ISR/RTOS timing, safe DMA ownership, and measured power budgets.