# Benchmark Pipeline State Snapshot
- Generated: 2026-03-04T06:10:55+00:00
- Scope: `ai_research/benchmarks` and `ai_research/benchmarks/runs`

## 1) Actual file paths for verifier / delta / normalize scripts
- `ai_research/benchmarks/generate_benchmark_delta.py`
- `ai_research/benchmarks/generate_benchmark_delta.sh`
- `ai_research/benchmarks/normalize_benchmark_summary.py`
- `ai_research/benchmarks/tests/test_normalize_benchmark_summary.py`
- `ai_research/benchmarks/verify_benchmark_artifacts.sh`

## 2) Current run artifact layout under `runs/`
- Root entries: `2026-03-04`, `2026-03-04-baseline-refresh`, `2026-03-06`, `2026-03-07-rebenchmark-delta`, `LATEST_RUN.txt`, `latest`, `latest_summary.json`

### 2026-03-04/
- `runs/2026-03-04/agi_repo_main_pycompile.stderr.log`
- `runs/2026-03-04/agi_repo_main_pycompile.stdout.log`
- `runs/2026-03-04/mk2_test_imports_pycompile.stderr.log`
- `runs/2026-03-04/mk2_test_imports_pycompile.stdout.log`
- `runs/2026-03-04/qapla_health_pycompile.stderr.log`
- `runs/2026-03-04/qapla_health_pycompile.stdout.log`
- `runs/2026-03-04/smoke.stderr.log`
- `runs/2026-03-04/smoke.stdout.log`
- `runs/2026-03-04/smoke_fail.stderr.log`
- `runs/2026-03-04/smoke_fail.stdout.log`
- `runs/2026-03-04/smoke_pass.stderr.log`
- `runs/2026-03-04/smoke_pass.stdout.log`
- `runs/2026-03-04/summary.json`

### 2026-03-04-baseline-refresh/
- `runs/2026-03-04-baseline-refresh/agi_repo_main_pycompile.stderr.log`
- `runs/2026-03-04-baseline-refresh/agi_repo_main_pycompile.stdout.log`
- `runs/2026-03-04-baseline-refresh/mk2_test_imports_pycompile.stderr.log`
- `runs/2026-03-04-baseline-refresh/mk2_test_imports_pycompile.stdout.log`
- `runs/2026-03-04-baseline-refresh/qapla_health_pycompile.stderr.log`
- `runs/2026-03-04-baseline-refresh/qapla_health_pycompile.stdout.log`
- `runs/2026-03-04-baseline-refresh/summary.json`

### 2026-03-06/
- `runs/2026-03-06/agi_repo_main_pycompile.stderr.log`
- `runs/2026-03-06/agi_repo_main_pycompile.stdout.log`
- `runs/2026-03-06/mk2_test_imports_pycompile.stderr.log`
- `runs/2026-03-06/mk2_test_imports_pycompile.stdout.log`
- `runs/2026-03-06/qapla_health_pycompile.stderr.log`
- `runs/2026-03-06/qapla_health_pycompile.stdout.log`
- `runs/2026-03-06/summary.json`

### 2026-03-07-rebenchmark-delta/
- `runs/2026-03-07-rebenchmark-delta/agi_repo_main_pycompile.stderr.log`
- `runs/2026-03-07-rebenchmark-delta/agi_repo_main_pycompile.stdout.log`
- `runs/2026-03-07-rebenchmark-delta/mk2_test_imports_pycompile.stderr.log`
- `runs/2026-03-07-rebenchmark-delta/mk2_test_imports_pycompile.stdout.log`
- `runs/2026-03-07-rebenchmark-delta/qapla_health_pycompile.stderr.log`
- `runs/2026-03-07-rebenchmark-delta/qapla_health_pycompile.stdout.log`
- `runs/2026-03-07-rebenchmark-delta/summary.json`

### LATEST_RUN.txt/ *(non-dir entry)*
- *(empty)*

### latest/
- `runs/latest/agi_repo_main_pycompile.stderr.log`
- `runs/latest/agi_repo_main_pycompile.stdout.log`
- `runs/latest/mk2_test_imports_pycompile.stderr.log`
- `runs/latest/mk2_test_imports_pycompile.stdout.log`
- `runs/latest/qapla_health_pycompile.stderr.log`
- `runs/latest/qapla_health_pycompile.stdout.log`
- `runs/latest/summary.json`

### latest_summary.json/ *(non-dir entry)*
- *(empty)*

## 3) Exact schema keys found in latest `summary.json`
- Latest summary file used: `ai_research/benchmarks/runs/latest/summary.json`
- mtime UTC: `2026-03-04T05:58:51+00:00`
- `benchmarks`
- `generated_at`
- `manifest`
- `root_dir`
- `run_id`
- `summary`

## 4) Gap table: acceptance-expected path -> real path or missing status
| Expected path | Status | Mapping / note |
|---|---|---|
| `ai_research/run_benchmarks.sh` | present | present at `ai_research/run_benchmarks.sh` |
| `ai_research/benchmarks/run_benchmarks.sh` | missing | missing at this path; control entrypoint exists at `ai_research/run_benchmarks.sh` |
| `ai_research/benchmarks/verify_benchmark_artifacts.sh` | present | present at `ai_research/benchmarks/verify_benchmark_artifacts.sh` |
| `ai_research/benchmarks/generate_benchmark_delta.sh` | present | present at `ai_research/benchmarks/generate_benchmark_delta.sh` |
| `ai_research/benchmarks/generate_benchmark_delta.py` | present | present at `ai_research/benchmarks/generate_benchmark_delta.py` |
| `ai_research/benchmarks/normalize_benchmark_summary.py` | present | present at `ai_research/benchmarks/normalize_benchmark_summary.py` |

## 5) Notes
- This is a concrete inventory snapshot for acceptance-path mismatch triage and pipeline reproducibility checks.