Blame for environment/README.md

# binary-reasoning-v1 — open sample

**Open sample: 100 of 10,251 verified reverse-engineering tasks**, plus all 29 agentic
flag-hunt challenge binaries with sources. The full pack (10,251 items, same schema) is a
paid product: complete-pack licensing with provenance, updates and grading material —
ask (samueljanttila@gmail.com), datasheet below.

Every item was produced by static analysis through our in-house analysis pipeline (Ghidra, binutils) and machine-checked:
answers are derived from the artifact itself and never leak into the input
(audited 0/10,251). Per-item provenance: binary path, sha256_16, address, arch.

## What is in this sample (100 items, selected by published script)
`select_sample.py` in this package reproduces the exact draw (seed 20261029, family quotas,
all minority-architecture items included). The sample includes gold answers so you can check
our graders before you buy.

| family | in full pack | in sample | grading |
|---|---|---|---|
| func_id_exact (identify stripped/internal function from disassembly) | 9,073 | 40 (all non-x86-64 items included: AArch64 11, RISC-V 3, PPC64 3, MIPS 2, i386 1, ARM 1; rest x86-64) | deterministic symbol match (alias-tolerant) |
| func_id_mc (multiple-choice identification) | 1,036 | 12 | deterministic option match |
| agentic_flag_hunt (find argv[1] that prints OK; binaries in `bins/`) | 29 | 29 (whole family — this is the public benchmark) | exact flag / stdout + exit-code probe |
| func_summary (summarize scrubbed function; rubric judge) | 53 | 6 | LLM judge + visible-constants rubric |
| crossarch_equivalence (same function across two ISAs?) | 25 | 5 | deterministic stance+symbol |
| library_ident (SBOM-style name+version from artifacts) | 14 | 3 | exact substring |
| arch/isa_semantics_qa (instruction semantics) | 14 | 3 | rubric judge |
| pe_import_reasoning / pe_structure_qa | 7 | 2 | rubric judge |

Sources: Ubuntu x86-64 system libraries (libcrypto, zstd, sqlite3, libxml2, zlib, libpng),
Windows PE samples, Debian production cross-toolchain glibc (AArch64/RISC-V/MIPS/PPC),
Android/Termux production aarch64 libs, and labeled cross-compiled targets.
The 29 agentic challenges are also packaged in Harbor task format on the HF dataset repo —
built + scored in real Docker containers: 29/29 (gold → reward 1, wrong answer → 0).

## Measured difficulty (honest)
- Agentic 29: Claude Opus 5.5 (tools: bash+read) solved 29/29 envs, 145/145 attempts, byte-exact
  vs gold in a non-isolated host run — that tier is warm-up level for a frontier agent; full table in EVAL_RESULTS-passk.md.
- Open-weight baseline (local quantized model, single-shot): 24/39 = 62% on a cross-arch + library-identification slice (n=39, 3-bit quantized, temperature 0). On the separate private hard set (99 items, not in this pack), a separate local 3-bit-quantized open-weight run (text-only prompt with no binary/file access, no tools, single-turn, temperature 0.7, max_tokens=600, whole-reply normalized matching) scored 14/99 at pass@8; that is headroom evidence, not a published band table or proof that each task sits in the RL mid-band. A fixed 20-item agentic spot check found the sampled command-output/static-analysis families practice-level for a frontier agent with shell tools (40/40 attempts; non-isolated host run, upper bound), so those families are being hardened before pilot use. We calibrate per buyer model.

Datasheet + full evaluation results: https://huggingface.co/datasets/S3RL4/binary-reasoning-v1
Canary for contamination checks: RE-Lab-brv-2026-10-29-8c1e4f7a

## Usage (verifiers v1)
```bash
uv run vf-eval binary_reasoning_v1 --select.limit 20
```
Or load in code:
```python
from verifiers import load_environment
env = load_environment("binary-reasoning-v1")   # 100-task open sample
```