SlimeCOBOL — COBOL → Java/Rust bit-exact transpiler
SKU structure (announced 2026-05-20)
SlimeCOBOL is now an umbrella of 5 SKUs
COBOL dialects, syntax, and preprocessor needs vary widely by lineage. SlimeCOBOL is organised as lineage-specific SKUs to provide narrow, hardened scope claims. Only the S1 front-end (tokenizer / preprocessor) is lineage-specific; the Slot IR + S2-S9 + 5 target-language emitters are fully shared (common backend). The Slot IR Ext32 dialect field was expanded from 4-bit to 8-bit (max 256 dialects) on 2026-05-20, providing headroom for all current and future SKUs.
>>source free + *> + COPY REPLACING. ACAS 3.01.07 production (~100K LOC) = 137/137 × 5 = 685/685 = 100.000% measured (2026-05-20).$SET / $IF / $ELSE directives. Release date TBD — early engagements accepted.“Narrow scope, hard claims”: every SKU stands on its own measured numbers; no SKU's claim is allowed to overstate beyond its scope. Adding a new SKU only requires one new S1 front-end — the backend (Slot IR + S2-S9 + emitters) is shared (the same architecture used by SlimeELF-rev / SlimePE-rev). JP Patent App. 2026-046620 (“Core64 + Ext32 fixed-bit Slot IR”) textual claim is unchanged.
An audit-grade COBOL → Java / Rust transpiler. Not a single cent drifts.
Lifts the decades-old legacy code of banks, pension systems, healthcare, and public-sector platforms into modern languages — bit-exact, down to the last bit.
- The transpiled code round-trips back token-for-token — conversion is provably reversible.
- 14 COBOL dialects auto-detected (IBM, Burroughs, Unisys, Hitachi OpenTP1, etc. — identifier-marker classification).
- Audited end-to-end against the NIST CCVS85 + Portfolio Mgmt suite (501 programs) → 501 / 501 = 100.000% program-level bit-exact, full audit, no sampling. As of 2026-05-20, also 4,595 / 4,595 byte-exact (100.000%) across 11 external corpora (919 files × 5 target languages, including ACAS 3.01.07 production system (~100K LOC GnuCOBOL)).
We make "does the transpiled code really behave the same?" a question your auditor can re-verify after the fact.
A deterministic legacy-modernization transpiler — combining bit-exact round-trip and hash-chain audit (the world's first to our knowledge).
2026-05-19 milestone — 10-corpus external WASI benchmark
SlimeNENC v1.4 Subset A was run through a 127.0.0.1-only WASI service (wasmtime 40.0.0) against 10 external COBOL corpora independent of our internal validation set. For every input, the HTTP response was verified to be SHA-256 byte-exact with the corresponding ELF-direct emit result (2026-05-19).
| Corpus | Source | Files | 5-lang tests | byte-exact PASS |
|---|---|---|---|---|
| NIST CCVS85 (FIPS 21-3) | US federal COBOL compiler certification suite | 459 | 2,295 | 2,295 / 2,295 = 100.000% |
| AWS CardDemo | aws-samples/aws-mainframe-modernization-carddemo (Apache 2.0; CICS/VSAM/JCL/RACF) | 44 | 220 | 220 / 220 = 100.000% |
| IBM CICS Banking Sample (CBSA) | cicsdev/cics-banking-sample-application-cbsa (EPL 2.0) | 29 | 145 | 145 / 145 = 100.000% |
| Rocket Software BankDemo | RocketSoftwareCOBOLandMainframe/BankDemo (commercial demo) | 81 | 405 | 405 / 405 = 100.000% |
| sentientsergio investment portfolio | sentientsergio/COBOL-Legacy-Benchmark-Suite (CICS/DB2/VSAM/Report Writer) | 42 | 210 | 210 / 210 = 100.000% |
| defects-suite (real-world bugs) | PhaseChangeSoftware/cobol-defects-suite (MIT) | 53 | 265 | 265 / 265 = 100.000% |
| DOGECICS (EXEC CICS + pseudo-conv + commarea + VSAM) | mainframed/DOGECICS (Apache 2.0) | 5 | 25 | 25 / 25 = 100.000% |
| writ3it / neopragma / michelou (OSS teaching corpora) | REDEFINES / ODO / PERFORM / Rosetta Code-style | 69 | 345 | 345 / 345 = 100.000% |
| ★ ACAS 3.01.07 production (Applewood accounting, ~100K LOC GnuCOBOL lineage) | SourceForge GPL, UK commercial since 1976, IRS / sales / purchase / general / stock / common modules | 137 | 685 | 685 / 685 = 100.000% |
| Total | 11 external corpora (one is a single ~100K LOC production system) | 919 | 4,595 | 4,595 / 4,595 = 100.000% |
Third-party reproducibility command:
wasmtime run --dir=. java_emit.wasm < input.cbl | sha256sum
(same procedure for Rust / C# / Kotlin / Go). Across all 4,595 tests
(incl. ACAS added 2026-05-20 via a S1 patch adding >>source free
directive support, opening the GnuCOBOL lineage), the HTTP service emit result
inside the WASI sandbox and the direct ELF emit result agreed on
every byte (SHA-256 identical). The April-23 dialect-verify
18/18 × 5 languages = 90/90 bit-exact cells are simultaneously preserved (regression 0).
Headline numbers
18 cases × 5 languages × 2 dialects
convert → back to source language → tokens match
IBM Enterprise, Burroughs/Unisys, Micro Focus, GnuCOBOL, Hitachi OpenTP1, etc. — identifier-marker classification.
※ dialect-transpile (actual conversion) coverage centers on GnuCOBOL / opensource-cobol; Hitachi etc. are at the fingerprint stage.
NIST CCVS85 + Portfolio Mgmt, full audit, no sampling
919 files × 5 target languages, including ACAS 3.01.07 (~100K LOC production) = 137 standalone × 5 (achieved 2026-05-20)
per program, p95 = 145 ms, scales O(N)
Validation corpus — the largest publicly-disclosed COBOL benchmark we are aware of in Japan
Note: the bit-exact figures are measured on the NIST CCVS85 + Portfolio Management subset (501 programs): program level is 501 / 501 = 100.000% (full audit, no sampling). As of 2026-05-20, an additional 4,595 / 4,595 byte-exact (100.000%) was measured across 11 external corpora (919 files × 5 target languages, including ACAS 3.01.07 production, ~100K LOC GnuCOBOL lineage) via a 127.0.0.1-only WASI service. The 2,876 programs / 1.15M LOC figures below describe the total validation corpus; their entirety has not been bit-exact measured (the additional volume is used for dialect verification, sales demos, and silent-pass true-value measurement).
NIST + AWS + IBM + 4 banking suites + Unisys + Burroughs and more
the largest publicly-disclosed COBOL benchmark we are aware of in Japan
NIST / AWS / IBM CBSA / db2-samples / Rocket / Defects Suite / UUSIG / etc.
Major sources include:
- NIST CCVS85 459 progs / 346K LOC — the US Federal compiler validation suite (FIPS 21-3)
- AWS Mainframe Modernization carddemo 106 progs / 41K LOC — Apache 2.0, official AWS demo
- IBM CBSA (CICS Banking) 66 progs / 27K LOC — EPL 2.0, IBM’s official banking sample
- IBM db2-samples cobol_mf 22 progs / 3.5K LOC — Apache 2.0, DB2 EXEC SQL
- Defects Suite (OpenCBS) 43 progs — MIT, real-world defects benchmark
- Rocket BankDemo 164 progs / 35K LOC — commercial banking demo
- VAULT-CBS 73 progs / 32K LOC — production-grade core banking
- UUSIG Public Domain archive of Unisys Users Group (Burroughs MCP)
- Burroughs B5500-software 13 progs — actual 1960s mainframe sources
Comparison vs. vanilla COBOL compilers
When the same corpus is syntax-checked with GnuCOBOL 3.1.2, the value gap is quantitative:
| Target / processor | Acceptance |
|---|---|
GnuCOBOL -std=cobol2014 (pure ISO) on Defects Suite | 35 % |
GnuCOBOL -std=ibm / -std=mf on Defects Suite | 75 % |
| GnuCOBOL on enterprise COBOL with EXEC CICS / EXEC SQL (CBSA / db2-samples / carddemo) | 0–6 % (preprocessor missing → rejected) |
| SlimeNENC (same domain, full audit on 501 NIST + Portfolio) | 100.000 % (501 / 501 program level) + external 10-corpus 3,910 / 3,910 byte-exact |
Observation: vanilla GnuCOBOL accepts only 0–6 % of enterprise COBOL with EXEC CICS / EXEC SQL because no preprocessor is available. SlimeNENC bit-exact transpiles the same domain at 501 / 501 = 100.000 % program level (full audit, no sampling). On top of that, 2026-05-20 measurement shows 4,595 / 4,595 byte-exact (100.000 %) across 11 external corpora (919 files × 5 target languages, including ACAS 3.01.07 production, ~100K LOC GnuCOBOL lineage), reproducible inside a WASI sandbox by third parties. The three-tier value staircase — ISO 35 % / IBM-MF 75 % / SlimeNENC 100.000 % — is presented as a chain of measured values.
Supported target languages
A single Slot IR (intermediate representation) feeds a per-language Stage 6 emitter. Stages S1–S5 and S7–S8 are language-agnostic and fully shared; adding a new target language requires only one new emitter file (Rust, Go, C#, and Kotlin emitters were each ported in roughly 30 minutes).
For Java specifically, we have validated cross-version transpilation across 8 ↔ 17 ↔ 21 (bytecode major versions 52 / 61 / 65), producing physical evidence of JVM compatibility across a 13-year version gap.
Pipeline (9 stages)
Each stage corresponds to an independent claim in the patent, and each behaves deterministically. On failure, Mini-PSV (Stage 8) and the audit chain (Stage 7) preserve the evidence.
(one file per language)
shuffle detection 100%
Audit suitability
- Bit-exactFor identical input, the transpiled code's computed output matches the source down to the last bit. COBOL features such as COMP-3, SIZE ERROR, GIVING, and multi-target are tuned per dialect (14 values detected, 18/18 dialect-verified bit-exact runs primarily on IBM / Unisys / MF families).
- Round-tripRe-converting the output back to the original language produces a token-by-token match (90/90 cases). This is a formal certificate that conversion did not discard information, and it is something an auditor can verify back-and-forth.
- Audit chainEvery transformation step receives a tamper-evident hash chain. After the fact, "who, when, by which rule" is reconstructible and verifiable (FNV-1a-64 + direct struct hashing).
- Hallucination detectionA trigram language model plus Mini-PSV (bigram) catches 95.0% of mistranslations / hallucinated output. Silent-pass (overlooked) rate is below 1.1% on the CORE set.
- DeterminismSame input + same version → bit-identical SHA-256. Output does not drift across parallel or GPU execution, which makes regression testing meaningful.
- Build-time LLM onlyLLMs are used only at rule-construction time. The runtime (actual transpilation) is purely deterministic and rule-based. This is the basis of our claim that "100.000% is a measured count (501 / 501 + 4,595 / 4,595), not a probability."
Typical use cases
| Finance | Migrate regional-bank and insurer COBOL batches (interest accruals, maturity notifications, etc.) to Java/Kotlin platforms without losing a single yen of interest. The 14-dialect auto-detect handles legacy IBM and Unisys implementations from the same engine. |
|---|---|
| Pension / Healthcare | For computations subject to long-term retention obligations, Round-trip + Audit chain proves that pre- and post-conversion outputs match exactly. Auditors can independently re-verify "no numerical drift introduced by conversion." |
| Public sector | Modernize 20+ year-old MUMPS / COBOL assets via a domestic transpiler, in incremental phases. Eliminates the indeterminacy of LLM-driven translation by replacing it with a formally guaranteed transformation. |
| Enterprise IT modernization | For Java 8 → 17 → 21 cross-version migration, we chain bytecode major 52 → 61 → 65 with bit-exact guarantees throughout. "Behavioral invariance" is demonstrable without an exhaustive regression suite. |
Five major mainframe vendors covered by dialect fingerprints
SlimeNENC’s dialect-detect engine carries identifier-marker fingerprints for the three major Japanese mainframe vendors plus two key US-origin platforms. For Coast Guard, municipal, regional-bank, international payment, and SMB-ERP engagements where multiple vendors coexist, “one engine that identifies all of them” is a clear differentiator.
Japanese mainframe vendors (3)
| Vendor | Primary markers | Calibration state |
|---|---|---|
| Hitachi (OpenTP1 / VOS3) | CBLEELOG / CBLEETRN / CBLEEMEM / CBLEERPC / CBLEEDBQ / CBLEETIM CBLEE\* API family, ADDRESSED BY clause, KFSB*****-{E,I,W} message-ID convention, P-1M64 product code | ✅ Calibrated against actual samples (OpenTP1 manual, 8 UAPs) |
| Fujitsu (NetCOBOL / GS21 / PowerCOBOL) | PFD-V*L* product code, AIM transaction base, Symfoware RDB, JEF kanji code, CALL 'PFM*' performance API | △ Public-spec based, calibration via PoC in progress |
| NEC (ACOS-4 / ACOS-6) | ACOS-4 / ACOS-6 OS, NCRP transaction, ADBS / RIQS II RDB, NX7700 / Express5800 hardware, S2900 / S3900 | △ Public-spec based, calibration via PoC in progress |
US-origin platforms (2) — international payment, SMB-ERP reach
| Vendor | Primary markers | Calibration state |
|---|---|---|
| HP NonStop / Tandem (Visa / Mastercard payment cores) | GUARDIAN OS, Enscribe file system, PATHWAY / PATHCOM transaction, IPCF IPC, ENTER TAL '...' clause, NOWAITED / WAITED I/O, S/H/J/NB-Series hardware, SQL/MP / SQL/MX | △ Public-spec based, awaiting US PoC for calibration |
| IBM i / AS/400 ILE (SMB ERP / public-sector residual) | CRTBNDCBL / CRTCBLPGM CL command, DDS / DSPF integration, CALL PROCEDURE, CALL 'Q*' system calls, QGPL / QSYS library, RTNVAL modifier, COBOL/400 (legacy name) | △ Initial validation against martinfx-cobol (28 files, GPL) |
Note: the above describes dialect-detect (identifier-marker classification). dialect-transpile (actual conversion) coverage centers on GnuCOBOL / opensource-cobol / IBM z/OS / MicroFocus / Burroughs; the five vendor families above are at the fingerprint stage, with conversion PoCs proceeding via customer-supplied code. Please contact us for details.
Subset structure
SlimeNENC consists of two subsets:
| Subset A | Cross-language transpilation (this page) Bit-exact migration to a different target language. The 90/90 evidence on this page belongs here. |
|---|---|
| Subset B | PSDP (Phase-Synchronous Deterministic Parallelism) Bit-exact within-language parallelization. A sequential Java batch becomes a parallel Java batch with no change in output. PSDP detail page → |
| SlimeJCL (companion) | JCL → Spring Batch transpiler (separate product, separate license) SlimeNENC-user-only distribution, targeted GA 2026-Q3. Bit-exactly carries the JCL counterpart of COBOL batches into Spring Batch, completing the migration project in a single toolchain. SlimeJCL detail page → |
Technical specifications
| Patent | JP Patent App. 2026-046620 (SlimeNENC family; one application covering Subset A cross-language transpilation and Subset B PSDP) |
|---|---|
| Claims mapped | Stage 1 (claim 1f) / Stages 2–3 (claim 3) / Stage 4 (claim 6) / Stage 5 (claims 2, 7, 8) / Stage 7 (claim 9) / Per-dialect Slot 14b/c/d |
| Standard tests | NIST CCVS85 501 programs / Internal 18 cases × 5 languages × 2 dialects = 90 cells |
| Performance | median 62 ms / p95 145 ms / WASM build within 2× overhead / O(N) scaling |
| Dialect coverage | dialect-detect (identifier-marker): 14 values (IBM Enterprise / Micro Focus / GnuCOBOL / Burroughs(Unisys) / Hitachi OpenTP1 etc.), dialect-verify bit-exact 18/18 (primarily on IBM/Unisys/MF families). dialect-transpile (actual conversion) coverage centers on GnuCOBOL / opensource-cobol; Hitachi etc. at fingerprint stage. |
| License model | WASM converter is licensed; converted output is unlicensed. Ed25519 3-hop activation. |
Related documentation
Detailed materials are available on request:
- Technical overviewSlimeNENC Technical Overview (A5, 21 pages, available as PDF / Markdown / HTML)
- PSDP paperPSDP Paper JP v5d (Subset B, 910 KB PDF)
- Patent specificationsJP Patent App. 2026-046620 (SlimeNENC family; one application covering Subset A and Subset B PSDP)
- Reference implementations5 languages × 23 converters (with 90/90 bit-exact validation)
- Benchmarks501-program × 8-stage performance profile (median / p95 / scaling)