Download .claude/skills-en/implementation-approach/SKILL.md from the publisher, or copy the path and open the repository. PlugMCP lists the skill and does not host the file.
Where does implementation-approach live?
The publisher keeps it at https://github.com/shinpr/ai-coding-project-boilerplate.
Which agents can use implementation-approach?
Claude, Cursor, Codex, and Grok Build can read a SKILL.md you add from the publisher's repository. Grok Bot installs it when you send the skill's GitHub folder link in a chat.
From the publisher
SKILL.md
name
implementation-approach
description
Selects implementation strategy (vertical slice, horizontal, or hybrid) with risk assessment. Use when planning feature implementation.
Phase 1: Decision-Sufficient Current State Analysis
Core Question: "What does the existing implementation look like?"
Analysis Framework
Architecture Analysis: Responsibility separation, data flow, dependencies, technical debt
Implementation Quality Assessment: Code quality, test coverage, performance, security
Historical Context Understanding: Current form rationale, past decision validity, constraint changes, requirement evolution
Meta-cognitive Question List
What is the true responsibility of this implementation?
Which parts are business essence and which derive from technical constraints?
What dependencies or implicit preconditions are unclear from the code?
What benefits and constraints does the current design bring?
Stop when another current-state fact cannot change responsibility, reuse, option validity, total complexity, a contract, or verification.
Completion evidence: inspected paths, observed architecture/data-flow facts, known constraints, inferred historical rationale labeled as inferred, and unknowns that could change strategy selection.
Transition: proceed when every strategy-relevant claim is observed, explicitly inferred with evidence, or recorded as unknown.
Phase 2: Design Convergence
Core Question: "What is the smallest design that delivers the current required outcome, and what evidence forces each addition beyond it?"
Complete these steps in order before exploring implementation strategies:
Existing-Surface Baseline: Form the simplest end-to-end path that delivers the current outcome through existing responsibilities. Explicit requirements and accepted decisions are binding; suggested mechanisms remain candidates.
Evidence Check: Test that path against current requirements, verified constraints, observed in-scope problems, and evidence-backed material risks. Keep only the unmet conditions that can change the selected design. Where the requirements leave a quality level unset, such as performance, freshness, consistency, availability, or protection beyond the required trust boundary, test against the lowest level that still delivers the outcome and record it as the selected level.
Targeted Comparison: For each unmet condition, test reuse, derivation from existing data, on-demand computation, or responsibility at the current caller or boundary before adding design surface. Compare viable choices by total complexity across the dimensions that materially differ: user decisions, settings, modes, concepts, outputs, persistent state, implementation paths, UX, runtime, implementation, testing, documentation, and maintenance. Select the lowest-total-complexity choice that satisfies the condition.
Subtraction Check: Remove each proposed addition and re-test its governing condition. Retain it only when the outcome, a required boundary, or necessary proof becomes unmet at its origin (a requirement, an accepted decision, or a repository rule or contract); a criterion, invariant, or constraint that a workflow artifact derived from that origin is narrowed to what the origin requires instead.
Candidate paths and rejected additions remain active analysis. The durable output is the Selected Design: the complete chosen path plus evidence for each added design surface and the condition that fails when it is removed. Only an accepted ADR may retain alternatives as decision history. During implementation, use the same convergence check without producing a separate artifact.
Completion evidence: one complete Selected Design; every added design surface names its current evidence, why lower-surface resolutions fail, and its subtraction result.
Transition: proceed when every supporting claim is observed, explicitly inferred with evidence, or recorded as unknown; route an unknown that blocks the next step as an exact evidence prerequisite. User interaction is required only when the unknown requires changing the outcome, desired-future requirements, or non-goals, or authorizing an irreversible action.
Phase 3: Strategy Exploration and Creation
Core Question: "When determining before -> after, what implementation patterns or strategies should be referenced?"
Strategy Discovery Process
Direct Strategy: The smallest repository-supported strategy that delivers the Selected Design within the hard constraints
Research Order: repository patterns first; then official documentation for the resolved dependency version; then maintained OSS implementations; use literature/blogs only as supplements labeled non-authoritative
Reference Strategy Patterns
Use a pattern only when its named migration or dependency problem exists in the repository.
Legacy Handling Strategies:
Strangler Pattern: Gradual migration through phased replacement
Facade Pattern: Complexity hiding through unified interface
Adapter Pattern: Bridge with existing systems
New Development Strategies:
Feature-driven Development: Vertical implementation prioritizing user value
Foundation-driven Development: Foundation-first construction prioritizing stability
Risk-driven Development: Prioritize addressing maximum risk elements
Integration/Migration Strategies:
Proxy Pattern: Transparent feature extension
Decorator Pattern: Phased enhancement of existing features
Bridge Pattern: Flexibility through abstraction
Completion evidence: the direct strategy mapped to the observed constraints it satisfies and the constraints it leaves unresolved, plus an alternative only when it would materially change risk, compatibility, rollout, or the early verification point.
Transition: proceed when the direct strategy and any retained alternative are comparable against the same constraint set.
Phase 4: Risk Assessment and Control
Core Question: "What risks arise when applying this to the existing implementation, and which control measurably reduces likelihood or impact while preserving verification and rollback?"
Evaluate only risk categories for which current evidence can change the strategy, public contract, rollout, rollback, or verification boundary.
Risk Analysis Matrix
Technical Risks: System impact, data consistency, performance degradation, integration complexity
Operational Risks: Service availability, deployment downtime, process changes, rollback procedures
Completion evidence: each evidenced material risk has likelihood/impact evidence and one preventive or containment control.
Transition: proceed when every high-impact risk has either a control or a blocking escalation.
Phase 5: Constraint Compatibility Verification
Core Question: "What are this project's constraints?"
Check only constraints evidenced by the governing requirements, repository, external contracts, or current environment that can change the selected strategy or verification boundary.
Completion evidence: each constraint is observed, inferred, or unknown; every unknown that can invalidate a candidate names the exact evidence prerequisite.
Transition: proceed when remaining unknowns cannot change the valid candidate set, or the user resolves them.
Phase 6: Implementation Approach Decision
Select the approach that satisfies all hard constraints and current requirements with the lowest transition risk and smallest verification delay. Use lifecycle cost and implementation effort only as tiebreakers after requirement coverage, compatibility, and risk control are equal.
Vertical Slice (Feature-driven)
Characteristics: Vertical implementation across all layers by feature unit
Application Conditions: Low inter-feature dependencies, output in user-usable form, changes needed across all architecture layers
Verification Method: End-user value delivery at each feature completion
Horizontal Slice (Foundation-driven)
Characteristics: Phased construction by architecture layer
Application Conditions: Foundation system stability important, multiple features depend on common foundation, layer-by-layer verification effective
Verification Method: Integrated operation verification when all foundation layers complete
Hybrid (Creative Combination)
Characteristics: Flexible combination according to project characteristics
Application Conditions: Unclear requirements, need to change approach per phase, transition from prototyping to full implementation
Verification Method: Assign L1 when the phase produces end-user-operable behavior, L2 when it produces a testable internal behavior or contract, and L3 only when the phase produces build-time structure with no runnable behavior yet
For Hybrid, assign one explicit L1/L2/L3 verification level and observable completion result to every phase.
Completion evidence: one selected approach, its phase boundaries, integration points, and a verification result for every phase.
Transition: proceed to documentation when the selected approach covers every hard constraint and its risks have controls. Otherwise return to candidate exploration (Phase 3), or to Design Convergence (Phase 2) when a Phase 4-5 result changes the Selected Design or its evidence.
Phase 7: Decision Rationale Documentation
Return the following structure in the Design Doc or planning handoff:
Candidate approaches and rejection reasoning remain active analysis unless an accepted ADR owns them as decision history.
Completion evidence: the selected approach and every added design surface trace to an observed constraint, accepted inference, or resolved value-boundary decision.
Verification Level Definitions
Priority for completion verification of each task:
L1: Functional Operation Verification - Operates as an end-user feature (e.g., a user can execute a search and receive results)
L2: Test Operation Verification - Tests that prove the changed behavior pass (e.g., type definition tests)
Priority: L1 > L2 > L3 in order of verifiability importance
Integration Point Definitions
Define integration points according to selected strategy:
Strangler-based: When switching between old and new systems for each feature
Feature-driven: When users can actually use the feature
Foundation-driven: When all architecture layers are ready and E2E tests pass
Hybrid: When individual goals defined for each phase are achieved
Decision Gate Checklist
Phase 1 evidence exists before strategy selection
Phase 2 produces one complete Selected Design and every added design surface maps to current evidence, lower-surface insufficiency, and a subtraction result that leaves its origin unmet
Candidate generation includes combinations when no listed strategy satisfies all hard constraints
Every evidenced material risk has a control
Every hard constraint maps to the selected approach
Phase 7 output records the selection, total-complexity basis, and added design surface; alternatives appear only in an accepted ADR
When evidence required by a checked item is unknown, stop at that phase and report the exact repository evidence prerequisite. User interaction is required only when the unknown requires changing the outcome, desired-future requirements, or non-goals, or authorizing an irreversible action.
Guidelines for Meta-cognitive Execution
Leverage Known Patterns: Use a pattern when its problem and trade-off match the observed repository state
Evidence-Ordered Research: Use repository evidence, version-matched official documentation, maintained OSS examples, then supplementary secondary sources
Apply 5 Whys: Pursue root causes to grasp essence
Multi-perspective Evaluation: Complete the evidence and transition checks for Phases 1-5
Strategy Composition: Combine strategies when one strategy cannot satisfy all hard constraints
Decision Traceability: Map every selection reason to evidence in the Phase 7 output
