first-fluke Gap Analysis
Systematic tool for identifying discrepancies between current implementation/state and desired requirements/design. Used primarily in the PLAN and REFINE phases to ensure alignment.
Role
- Analyze current codebase or system state against requirements.
- Identify missing components, deviations, or technical debt.
- Provide actionable recommendations to bridge the identified gaps.
- Verify improved alignment after changes.
Core Rules
- Evidence-Based: All gaps must be supported by specific file paths, code snippets, or configuration values.
- Actionable: Each gap must have a clear remediation step (e.g., "Implement X", "Refactor Y").
- Scoped: Focus on the specific domain requested (e.g., Security, Architecture, Feature Parity).
- No Assumptions: Verify the current state using tools (
view_file,grep_search) before declaring a gap.
Workflow
Step 1: Definition of States
Define the "Desired State" (from requirements/docs) and "Current State" (from codebase/runtime).
Use read_file to fetch requirements and search_for_pattern/find_symbol to explore current code.
Step 2: Gap Identification
Compare the two states. Categorize gaps into:
- Missing: Feature exists in design but not in code.
- Deviation: Feature exists but behaves differently or has wrong signature.
- Extra: Code exists that is not in design (scope creep).
Step 3: Impact Analysis
Assess the impact of each gap (Blocking, High, Medium, Low). Determines priority for remediation.
Step 4: Remediation Plan
Generate a list of tasks to close the gaps. Output format should be compatible with Task Tracking (e.g., Markdown checklist or JSON).
Resources
- Protocol:
resources/execution-protocol.md(Detailed checklist and methodology) - Examples:
resources/examples.md(Common gap patterns)