docs: add Narsil-MCP competitive deep-dive with feature candidates

[carlos-alm]

Summary

• Add comprehensive competitive deep-dive comparing codegraph vs Narsil-MCP across 10 dimensions (parsing, graph model, search, security, queries, performance, install, MCP integration, developer productivity, ecosystem)
• Append Narsil-Inspired Feature Candidates section with 15 actionable features graded using the same tier system as BACKLOG.md

Feature candidates added

Tier 1 non-breaking (10): MCP tool presets (N1), AST-aware chunking (N2), code similarity search (N3), git blame & symbol history (N4), remote repo indexing (N5), config wizard (N6), Kotlin (N7), Swift (N8), Bash (N9), Scala (N10)

Tier 1 breaking (1): Export map per module (N11)

Tier 2 (2): Interactive HTML visualization (N12), multiple embedding backends (N13)

Tier 3 (2): OWASP pattern detection (N14), SBOM generation (N15)

Not adopted (10): Taint analysis, type inference, SPARQL/RDF, CCG standard, in-memory architecture, 90-tool surface, browser WASM, Forgemax, LSP integration, license scanning — all with FOUNDATION.md reasoning

Cross-references: BACKLOG IDs 7, 8, 10, 14 and Joern candidates J4, J5, J8, J9

Test plan

• Verify all internal links resolve (BACKLOG.md, COMPETITIVE_ANALYSIS.md, joern.md, FOUNDATION.md)
• Confirm no source code changes — docs only
• Review feature candidate grading against BACKLOG.md scoring guide

[greptile-apps]

Greptile Summary

Adds comprehensive competitive deep-dive for Narsil-MCP and appends feature candidates to both competitive analysis documents.

Key additions:

• narsil-mcp.md: 415-line analysis across 10 dimensions (parsing, graph model, search, security, queries, performance, install, MCP integration, developer productivity, ecosystem) with detailed comparison tables, principle-by-principle evaluation against FOUNDATION.md, and 15 Narsil-inspired feature candidates (N1-N15)
• joern.md: Appends 11 Joern-inspired feature candidates (J1-J11) using same BACKLOG.md tier system with proper cross-references

Grading consistency:

• All 26 features use consistent BACKLOG.md tier system (Tier 1 non-breaking, Tier 1 breaking, Tier 2, Tier 3, Not adopted)
• Proper cross-references to existing BACKLOG IDs (7, 8, 10, 14) and between documents (Joern J4/J5/J8/J9 ↔ Narsil N7/N8/N2)
• All internal links verified: BACKLOG.md, COMPETITIVE_ANALYSIS.md, FOUNDATION.md references resolve correctly

Documentation quality:

• Professional tone, well-structured tables, proper markdown formatting
• All @ symbols properly wrapped in backticks
• No duplicate feature IDs
• Follows established competitive analysis template

Confidence Score: 5/5

• Documentation-only PR with no code changes - safe to merge
• Perfect score because: (1) documentation-only changes with zero code impact, (2) all internal links and cross-references verified and correct, (3) consistent formatting and structure following established templates, (4) no duplicate IDs or broken references, (5) comprehensive test plan covers all validation points
• No files require special attention

Important Files Changed

Filename	Overview
generated/competitive/narsil-mcp.md	New comprehensive competitive analysis document with 415 lines covering 10 comparison dimensions and 15 feature candidates (N1-N15) graded using BACKLOG.md tier system
generated/competitive/joern.md	Added Joern-Inspired Feature Candidates section with 11 features (J1-J11) using same tier/grading system, includes cross-references to BACKLOG IDs 7 and 14

_{Last reviewed commit: 0faf02c}

[greptile-apps]

_{2 files reviewed, no comments}

_{Edit Code Review Agent Settings | Greptile}

[claude]

Claude finished @carlos-alm's task —— View job

---

Code Review: Documentation Analysis

Todo List:

• Read modified documentation file
• Verify internal link references
• Check feature candidate grading consistency
• Assess documentation quality and structure
• Provide final review recommendation

Root Cause Analysis

Problem Being Solved: Adding comprehensive competitive documentation comparing codegraph vs Narsil-MCP with 15 feature candidates graded using the BACKLOG.md tier system. This supports strategic planning by identifying valuable features from a close competitor.

Why This Approach: The PR adds structured competitive analysis alongside feature candidates with consistent grading, following the established template pattern from joern.md. The documentation provides strategic value for product direction.

Risk Assessment: Documentation-only change with zero code impact. Low risk.

✅ Documentation Quality Assessment

Strengths:

• Comprehensive analysis: 415 lines covering 10 dimensions (parsing, graph model, search, security, queries, performance, install, MCP integration, developer productivity, ecosystem)
• Consistent formatting: Professional tables, proper markdown, all @ symbols correctly wrapped in backticks
• Accurate cross-references: All internal links verified (BACKLOG.md, FOUNDATION.md, COMPETITIVE_ANALYSIS.md, joern.md)
• Grading integrity: All 15 features (N1-N15) use exact BACKLOG.md tier system with correct Zero-dep/Foundation-aligned/Problem-fit/Breaking assessments
• No duplicates: No duplicate feature IDs, clean ID sequence N1-N15
• Principle-based evaluation: Evaluates Narsil against FOUNDATION.md principles with specific scores (Codegraph 4, Narsil 0, Tie 4)

Technical accuracy verified:

• Feature candidate cross-references to existing BACKLOG IDs (7, 8, 10, 14) are accurate
• Joern candidate cross-references (J4↔N7, J5↔N8, J8, J9↔N2) are correct
• All markdown formatting follows established competitive analysis template

Final Recommendation

• Rating: ⭐⭐⭐⭐⭐ (5/5)
• Action: APPROVE
• Reasoning: This is exceptional documentation work with zero code impact. Perfect adherence to established grading system, comprehensive competitive analysis, verified cross-references, and professional structure. The 15 feature candidates provide genuine strategic value for product direction. No issues found.