Raw source exploration
Repository source access with normal file discovery and inspection.
Intended use
Baseline investigation when source bodies and tests are the only available context.
A controlled cross-language evaluation of raw source exploration, PViz-guided source reads, and bundle-only analysis.
Evaluated sessions
84
Tasks Per Repository
4
Repositories
7
Context strategies
3
The question
This evaluation compares normal source exploration, PViz-guided source verification, and bundle-only analysis. The goal is not to crown a universal winner; it is to identify which context strategy is appropriate for which class of repository-analysis task.
Raw source exploration
Repository source access with normal file discovery and inspection.
Intended use
Baseline investigation when source bodies and tests are the only available context.
PViz-assisted investigation
PViz structural bundle plus targeted source reads for verification.
Intended use
Source-grounded investigation with structural navigation, graph metrics, and scoped verification.
PViz Solo
PViz bundle only; no repository source reads.
Intended use
Architectural orientation, structural risk discovery, and explicitly bounded first-pass analysis.
Methodology
Assessment sequence
Define the task record
Write the prompt, task-family classification, and source-grounded ground-truth card before reviewing any condition answer.
Run three context conditions
Evaluate Raw source exploration, PViz-assisted investigation, and PViz Solo against the same task.
Assess independently
Review each condition against the shared ground-truth record rather than against the other condition outputs.
Score and preserve evidence
Score D1–D5, retain transcripts and metadata, then synthesize results by task family rather than as a universal winner ranking.
Scoring dimensions
Model Used
All 84 sessions were run using Claude Sonnet 4.6 via Claude Code. Conditions differed only in the context supplied to the agent — Raw, PViz-assisted, or bundle-only — with the model and harness held constant across all sessions.
Central finding
Behavioral-semantic questions require source-level verification. Mixed tasks benefit most from PViz-assisted work because structural metrics and source semantics are both material. Structural-native questions can often be oriented from the bundle alone, with uncertainty explicitly bounded.
| Task family | What it asks | Best-supported mode | Why |
|---|---|---|---|
| Behavioral / semantic | What happens, in what order, under which conditions? | Source-backed analysis | Method bodies, branches, assertions, and runtime semantics are decisive. |
| Mixed | What is structurally affected, and what behavior makes that matter? | PViz-assisted investigation | The bundle supplies structural metrics and orientation; source verifies behavioral detail. |
| Structural-native | What is connected, risky, cyclic, or widely depended on? | PViz Solo or PViz-assisted | The decisive facts are often directly represented in the structural bundle. |
Aggregate results
| Condition | D1 + D2 | Total D1–D5 | Interpretation |
|---|---|---|---|
| Raw source exploration | 192/196 (98.0%) | 407/420 (96.9%) | Near-ceiling source baseline across the benchmark. |
| PViz-assisted investigation | 196/196 (100.0%) | 400/420 (95.2%) | Matched or exceeded Raw on substantive scoring; total-score deductions were concentrated in evidence-efficiency overhead and known harness artifacts. |
| PViz Solo | 131/196 (66.8%) | 326/420 (77.6%) | Strong structural orientation with a consistent boundary on behavior-sensitive claims. |
Practical guidance
Baseline investigation when source bodies and tests are the only available context.
Source-grounded investigation with structural navigation, graph metrics, and scoped verification.
Architectural orientation, structural risk discovery, and explicitly bounded first-pass analysis.
Representative evidence
Example task packet
Each task packet contains everything needed to verify a result independently — the original prompt, a source-grounded ground-truth card, per-condition transcripts and metadata, and the final assessment with dimension scores.
Example
Go · Caddy
Mixed
PViz supplied risk, SCC, importer, and cycle context while using fewer tool calls than the Raw condition.
Condition records
Limits and interpretation
Scope set before evaluation
These boundaries were part of the evaluation design and define the claims the benchmark was intended to support.
Identified during evaluation
These findings emerged while running or scoring the benchmark and affect how specific results should be interpreted.
Audit trail
Evaluation evidence manifest
Index of the 28 task packets, 84 condition transcripts, metadata, and associated archive artifacts.
JSON
Cross-language summary
Aggregate scoring, task-family analysis, contribution patterns, anomalies, and methodology notes.
Go summary
Per-language findings, task-family breakdown, and anomaly notes for Caddy.
Java summary
Per-language findings, task-family breakdown, and anomaly notes for Micrometer.
Kotlin summary
Per-language findings, task-family breakdown, and anomaly notes for kotlinx.coroutines.
Python summary
Per-language findings, task-family breakdown, and anomaly notes for Textual.
Ruby summary
Per-language findings, task-family breakdown, and anomaly notes for Huginn.
Rust summary
Per-language findings, task-family breakdown, and anomaly notes for ripgrep.
TypeScript summary
Per-language findings, task-family breakdown, and anomaly notes for Vite.