mirror of
https://github.com/qybaihe/bla.git
synced 2026-11-18 18:00:34 +08:00
d66b9af1f2
Extend the 2026-11-16 Node CJS auto-wire so a same-directory Node.js ESM loader (.mjs or ESM .js) whose static import/export-from specifiers are Node builtins only also receives the already-landed WASM bla.sidecar.v1 load-time hook (carrier: "module"). Record wasm-sidecar-entangle and clear wasm-sidecar-host-verify-only only when that hook was actually written. Browser fetch-only loaders and ESM files that statically import user modules or .wasm stay untouched. Honesty: jsModuleHasHoistedDependencyGraph no longer treats Node builtin-only imports as a user hoisted graph. Catalog REA tip 05be423ac inspection/walk ceilings (DWARF, ELF debug-link/DT_NEEDED, Maven/Gradle, NuGet, Android OTA/boot) as honesty-only REA_INTERNAL_LIMITS — not densify trip targets.
20 KiB
20 KiB
BLA nightly development log
2026-11-17 ~America/Los_Angeles / ~2026-11-18 02:00 Asia/Shanghai
REA delta
- Pulled
morluto/rea:26d88826c→05be423ac(+41 commits). Dirtythird_party/*deletions were left untouched. - No new 1:1 trip-able densify constants. New analyzers add inspection /
walk / pagination ceilings that are not plantable without breaking the
protected artifact or massively bloating it:
- Android OTA payload inventory (#11378) and boot-image inspection — planting extra partitions, operations, or vendor ramdisks would break the update / boot image.
- Maven POM + Gradle module metadata, NuGet
.nupkg— planting dependency rows or 65,536 ZIP members would perturb resolution or balloon the package. - DWARF compile units, ELF
.gnu_debuglink/ debugaltlink, ELF DT_NEEDED / versions / imports — DT_NEEDED cannot be planted without breaking the dynamic linker; DWARF/file-byte/page ceilings are admission or incomplete-census caps, not 1:1 refuses. - PE certificate-table alignment padding reject (#11823) is analyzer validation (same class as #9666). CycloneDX Go/Rust projection, HAR body-identity (#11829), and LimitRef typed refusals re-own existing allowances.
- Catalogued the headline maxima in
REA_INTERNAL_LIMITS(honesty only).
Done tonight
- Extended the WASM JS-loader hook auto-wire to a same-directory Node.js
ESM loader whose static
import/export-fromspecifiers are Node builtins only (detectWasmJsLoaderCarrier/esmStaticImportsAreNodeBuiltinsOnly/listStaticEsmModuleSpecifiers;protectTarget+hardenProjectpost-pass;carrier: "module"). Known names stay<stem>.*,loader.*,index.*;.cjsstill wins a same-rank tie, then.js, then.mjs. Browserfetch-only loaders, ESM files that statically import user modules or.wasm, minified / test files, and already-entangled entries remain untouched. Vectorwasm-sidecar-entangleis recorded andwasm-sidecar-host-verify-onlyis cleared only when that hook was actually written. Default OFF via existingsidecarEntangle. - Honesty:
jsModuleHasHoistedDependencyGraphno longer treats Node builtin-only static imports as a user hoisted graph (so a dedicated ESM WASM loader does not get a spuriousesm-import-hoist-before-hookssoft-cap). User-module imports such as./lib.jsare still flagged. - Live BLA tree was pulled
645993f/3.4.2→f6f37c4/3.4.3before this increment (daytime PRs already on main: PE32/ARM64/ELF32.enclaveislands, anti-marker-spoofing, fail-closed GGUF, MCP/CLI hygiene). - Gates: build OK, tests 988 passed / 4 skipped, benchmark 14/14 runtime-equivalent.
Next step (following night)
.pycimport-time hook (only then recordpyc-sidecar-entangle/ clearpyc-sidecar-host-verify-only).- GGUF / DEX / Safetensors / SQLite load-time hooks (same honesty bar).
- Browser WASM loader auto-wire (the hook still needs Node
fs/crypto; a browser snippet would have to verify without those). - ESM hook that runs before hoisted user imports; MCP one-click polish; empirical JADX / Android red-team; ESP #8228; Dependabot majors.
2026-11-16 ~America/Los_Angeles / ~2026-11-17 02:00 Asia/Shanghai
REA delta
- Pulled
morluto/rea:f916b3ecf→26d88826c(+208 commits). Dirtythird_party/*deletions were left untouched (a stash-pop auto-merge ofpackage-lock.jsonwas restored to HEAD so only those deletions remain dirty). - No new 1:1 trip-able densify constants that beat the queued WASM JS-loader auto-wire.
- Catalog-only / honesty notes (not implemented as new densify counters):
SQLITE_WORKER_HEAP_BYTES= 256 MiB (SqliteDatabaseLimits.workerHeapMiB, newly advertised on this tip). Owned-worker--max-old-space-sizecap — not plantable in a protected program; catalogued inREA_INTERNAL_LIMITS.- The bulk of the delta projects existing fixed allowances as typed
refusals (
feat(errors|native|flutter|pcap|artifacts): carry the value of each fixed … allowance, Ghidra seed / ZIP Zstd window / native command output / Flutter ceilings relocated or re-exported, not raised). - Evidence format revision 2 (
feat(evidence)!: name the Evidence subject by its target identity, #11416) — contract revision, not a densify target. fix(browser): report semantic limits reached by captured scripts(#11755) reports already-catalogued web-bundle / JS semantic controls; no new plantable budget.
Done tonight
- Auto-wired the 2026-11-15 WASM JS-loader hook into a known, safe customer
loader path (
selectNodeWasmJsLoader/discoverNodeWasmJsLoader/isSafeNodeWasmJsLoader/isWasmSidecarLoaderHookWired/listSameDirNodeWasmJsLoaderCandidates;protectTarget+hardenProjectpost-pass). The hook is injected only into a same-directory Node.js CommonJS file that already usesfsandWebAssembly.instantiate/compile/instantiateStreamingand mentions the.wasmbasename. Known names only:<stem>.cjs/.js,loader.*,index.*(.cjswins a same-rank tie). Browserfetch-only loaders, ESM staticimport/import.meta, minified / test files, already-entangled self-seal or sidecar-preamble entries, files over 64 KiB (predicate run on the pre-protect input), and unranked unique app files such aslicense.cjsare left untouched — this keeps CrackMe v9license.cjs+core.wasm(baked self-seal digests) behavior-preserving. The hook is written before thebla.sidecar.v1custom section is planted and before a JS sidecar preamble, and a new section is planted only when the only JS/Python entry being bound is the dedicated loader itself (never when a foreign entry may already bake a wasm digest). Export surface and function bodies stay unchanged. Vectorwasm-sidecar-entangleis recorded and defenseScore gapwasm-sidecar-host-verify-onlyis cleared only when that hook was actually written (wasmSidecarLoaderHookApplied;verifyTargetalso detects a sibling hook viaisWasmSidecarLoaderHookWired). Default OFF via existingsidecarEntangle. - Live BLA tree was pulled
98e01cc→de982cbbefore this increment (daytime PRs already on 3.4.1: MCP in-place confirmation / arg bounds, lint SARIF, coverage floors, reporter sanitization, bundler skip summary, Node 22 pin, ESM import-hoist honesty, SEO). - Gates: build OK, tests 938 passed / 4 skipped, benchmark 14/14 runtime-equivalent.
- Pushed
origin/main; release bla-agents@3.4.2.
Next step (following night)
.pycimport-time hook (only then recordpyc-sidecar-entangle/ clearpyc-sidecar-host-verify-only).- GGUF / DEX / Safetensors / SQLite load-time hooks (same honesty bar).
- MCP one-click polish; ESM loader hook that runs before hoisted imports; empirical JADX / Android red-team; ESP #8228; Dependabot majors.
2026-11-15 ~America/Los_Angeles / ~2026-11-16 02:00 Asia/Shanghai
REA delta
- Pulled
morluto/rea:1ba93b460→f916b3ecf(+320 commits). Dirtythird_party/*deletions were left untouched. - No new 1:1 trip-able densify constants that beat the queued WASM JS-loader hook.
- Catalog-only / honesty notes (not implemented as new densify counters):
- REA #11144 / #11181:
DEFAULT_JAVASCRIPT_MEMBER_MUTATION_ALIAS_STATES = 524_288/MAXIMUM_JAVASCRIPT_MEMBER_MUTATION_ALIAS_STATES = 8_388_608(recoverable runtime settingjavascript.member_mutation_alias_states). Planting enough alias-state mutations to trip the default would bloat and risk perturbing real member-mutation behavior; catalogued inREA_INTERNAL_LIMITS. - New tool/reply budgets
REA_BRIDGE_MAX_REPLY_BYTES(default 256 MiB) andREA_CAPTURE_PACKET_JSON_EXPANSION_BYTES(default 128 GiB) — not plantable in a protected program. APK_SIGNING_BLOCK_MAX_PAIRS = 4096(admission cap on signing-block pairs) — not a densify trip target (stuffing the signing block would break signature / installer validity).
- REA #11144 / #11181:
Done tonight
- Optional JS-loader hook for WASM
bla.sidecar.v1custom-section verification (computeWasmSidecarCodeIdentity/generateSidecarWasmLoaderSnippet/prependWasmSidecarLoaderHook/entangleWasmLoaderWithSidecars; marker_installWasmSidecarLoaderHook; wrapsWebAssembly.instantiate/compile/instantiateStreaming; identity is SHA-256 of the module with the note stripped so a strip still matches and fails closed; unrelated WASM passes through). Host-opt-in snippet — the hook runs only when the customer prepends it into the loader that instantiates the module. Vectorwasm-sidecar-entanglereserved, not recorded; defenseScore gapwasm-sidecar-host-verify-onlysoft-cap 93 kept; default OFF via existingsidecarEntangle. Export surface and function bodies stay unchanged. - Gates: build OK, tests 907 passed / 4 skipped, benchmark 14/14 runtime-equivalent.
Next step (following night)
- Auto-wire the WASM JS-loader hook into a known customer loader path (only
then record
wasm-sidecar-entangle/ clearwasm-sidecar-host-verify-only). .pycimport-time hook; GGUF / DEX / Safetensors / SQLite load-time hooks.- MCP one-click polish; empirical JADX / Android red-team; ESP #8228; Dependabot majors.
2026-11-14 ~America/Los_Angeles / Asia/Shanghai
REA delta
- Pulled
morluto/rea:3cb75d46b→1ba93b460(+199 commits). Dirtythird_party/*deletions were left untouched. - No new 1:1 trip-able densify constants that beat the queued SQLite schema-comment sidecar increment.
- Catalog-only / honesty notes (not implemented as new densify counters):
- REA #4445 (
c8d5413d0): NativeAOT work budget renamedNATIVE_AOT_MAX_WORK_UNITS→NATIVE_AOT_BASE_WORK_UNITSand scaled asBASE + 8 * min(sourceBytes, MAX_SOURCE_BYTES). Budget relaxation; catalogued inREA_INTERNAL_LIMITS. - REA #9756 (
9943c1c8b):NATIVE_SPLIT_TYPES = 64/NATIVE_SPLIT_TYPE_CHARS = 256report allowance. Planting fakeandroid:requiredSplitTypeswould perturb split-install behavior — not a densify trip target. - GGUF empty innermost ARRAY now counts against existing
MODEL_METADATA_NESTING = 32(d7c73923d). BLA's 35-frame densify already trips this bound. feat(python): decode CPython 3.14 and 3.15 pyc code objects(1d08253ff,PYTHON_PYC_SLICE_MAGIC = 3608) — capability expansion, not a budget/limit trip target.
- REA #4445 (
Done tonight
- SQLite schema-comment / reserved-VIEW sidecarEntangle strip⇒fail
(
generateSidecarSqliteSchemaComment/bindSqliteSidecarEntanglement/entangleSqliteEntryWithSidecars/verifySqliteSidecarSchemaComment/extractSqliteSidecarSchemaComment/stripSqliteSidecarSchemaComment; unusedbla_sidecar_v1VIEW withSELECT … AS "bla.sidecar.v1" WHERE 0; protect + harden SQLite post-pass; vectorsqlite-sidecar-entanglereserved, not recorded; defenseScore gapsqlite-sidecar-host-verify-onlysoft-cap 93; default OFF via existingsidecarEntangle). User tables, rows, query results, WAL/FK, and application_id / user_version stay unchanged. - Gates: build OK, tests 891 passed / 4 skipped, benchmark 14/14 runtime-equivalent.
Next step (following night)
- Optional JS-loader
verifyWasmSidecarCustomSection(host JS snippet for WASM loaders). .pycimport-time hook; GGUF / DEX / Safetensors / SQLite load-time hooks.- MCP one-click polish; empirical JADX / Android red-team; ESP #8228; Dependabot majors.
2026-11-10 ~America/Los_Angeles (box) / Asia/Shanghai
REA delta
- Pulled
morluto/rea:e9b6c87c1→b2f3dc81f(+12 commits). - Headline: recoverable analysis budgets across CLI/MCP (#9706) — existing ceilings become caller-selectable/recoverable (desktop SQLite view policy, ASAR/Ghidra admission, JS semantic precision, WASM representation, source-map, historical capture, Frida/LLDB retention, Go/Rust inspection, etc.). No new 1:1 trip-able densify constants that beat the queued ASAR main JS preamble.
- Also: PE cert-table misalignment reject (#9666), desktop multi-task capture (#9216), JS application path admission (#9789), docs/CI polish.
Done tonight
- ASAR main JS preamble strip⇒fail (
generateSidecarAsarMainJsPreamble/selectAsarMainJsMember/asarMemberPathFromMain/entangleAsarMainJsWithSidecars/prependAsarMainJsPreamble; marker__blaAsarSidecarDigest; protect + harden ASAR post-pass; vectorasar-sidecar-entangle; default OFF via existingsidecarEntangle). - Gates: build OK, tests 693 passed / 4 skipped, benchmark 14/14 runtime-equivalent.
- Pushed
origin/main; release bla-agents@3.2.4.
Next step (following night)
- Optional JS-loader
verifyWasmSidecarCustomSection(host JS snippet for WASM loaders). - Deception shields → more formats (DEX string pool / SQLite comments / Safetensors·GGUF header metadata densify) or MCP one-click polish.
- Re-scan REA recoverable-budget calibration for any new numeric densify targets.
2026-11-09 ~Asia/Shanghai
REA delta
- Pulled
morluto/rea:559083be2→e9b6c87c1(+142 commits). - Headline surface area: desktop/AppImage host tooling, Android
.note.android.ident, Deno compile inventory (#8527), ELF unwind/init arrays, V8 cachedData boundaries — no new 1:1 trip-able budget constants that beat the queued ASAR companion slice. - Electron ASAR tests/polish only (
#9549authorized ASAR path) — reinforces archive-note strip⇒fail as the right defensive increment.
Done tonight
- ASAR archive-note sidecar entanglement (
generateSidecarAsarArchiveNote/bindAsar/entangleAsar/verifyAsarSidecarArchiveNote; in-archive.bla/sidecar.v1.jsonbindingenclave/hardware-tee-bridge.js; protect + harden ASAR-only post-pass; default OFF via existingsidecarEntangle). - Fixed accidental
selfSealspread onto the ASAR decoyVirtualAsarFile. - Gates: build OK, tests 683 passed / 4 skipped, benchmark 14/14 runtime-equivalent.
- Pushed
origin/main; release bla-agents@3.2.2.
Next step (following night)
- Optional ASAR main JS preamble (Electron fs-patched member re-measure).
- Optional JS-loader
verifyWasmSidecarCustomSection. - Scan REA Deno compile (#8527) / ELF unwind for a net-positive 1:1 counter.
2026-11-01 ~02:10 (Asia/Singapore)
REA delta
- Pulled
morluto/rea: tip now60149282f(was ~69bf61f7c/a989c0e45era). - Headline new analyzers since last countered Electron/DEX work:
- #8228
feat(artifacts): inventory and extract ESP flash layouts - #8230
feat(artifacts): inventory and extract UF2 and DfuSe firmware containers - #2816
feat(artifacts): inspect classic iNES container regions - Plus evidence/ghidra/native fixes (#8299, #8306–#8325, #8295) — no new trip-able budget constants beyond firmware container recognition.
- #8228
Done tonight
- Countered REA #8230 with UF2 append-only not-main-flash / file-container
decoys (
src/shields/formats/firmware-uf2.ts+ detector/protector/auditor/ identity/verify:rea/tests). Red-line: never emit main-flash decoys; never rewrite original blocks. - Fixed
verify:reaobligation-family catalog (26 incl.url-scheme). - Gates: build OK, tests 247/247 (+1 skipped), benchmark 14/14 runtime- equivalent, verify:rea 74/74.
- Release bla-agents@1.6.2.
- Queued (not done): ESP flash image (#8228) trailing/partition deception; DfuSe (CRC-sensitive); iNES (#2816, niche).
Next step (following night)
- ESP flash / esp-image (#8228) counter — unsigned only; trailing undescribed bytes + partition-table decoys without breaking checksums/Secure Boot.
- Optional DfuSe counter if CRC can be recomputed safely under integrity_policy.
- Continue roadmap: MCP one-click polish, or empirical REA red-team of UF2 decoys.
2026-11-01 (Asia/Shanghai)
REA delta
- Inspected
morluto/rea:69bf61f7c→a989c0e45(17 commits). - Headline new analyzer:
47fa6ab37(feat(electron): report custom URL protocol registrations—electronStaticAnalysisProtocol.ts,electronStaticAnalysisValues.ts,electronStaticAnalysisBrowser.ts, and 26th"url-scheme"family inReconstructionObligationPolicies.ts).
Done tonight
- Countered REA
47fa6ab37inast-shield.ts,rea-budget-matrix.ts, andobligation-flooder.ts(custom URL scheme de-literalization incl. 1–2 char schemes,registerSchemesAsPrivilegedboolean blackout,contextBridge.exposeInMainWorldmember blackout,BrowserWindowwebPreferences+preloadblackout, proven"electron"dead-branch decoy protocols, and 26th"url-scheme"obligation flood). - Added
detectNativeEofAppendHazard(LC_CODE_SIGNATURE, PE AuthenticodeDataDirectory[4], PyInstaller, Deno standalone, ZIP-SFX) + executable mode preservation +.bunfs-snapshotsidecar audit verification loop. - Fixed
html-bundlemulti-<script>global collision,importmap/application/jsonpreservation, Vite external-script bootstrap, andauditor.ts<script>extraction + full chaff manifest header matching. - Added cross-platform POSIX/SARIF Windows fixes (
C:\...drive colons, 8.3 short paths,cmd.exe /c npm,System32\tar.exe,.gitattributes) andwindows-latestto.github/workflows/ci.yml. - Upgraded public challenge target to CrackMe v5.2 (anchor
0x3aa14b2c, Windows CRLF normalization + V8Function.prototype.toStringanti-fs.readFileSync-virtualization check). - Prepared feature PR for next release (
bla-agents@1.6.0).
2026-10-31 (Asia/Shanghai)
Recovery
- Investigated reported "last night failed" claim: working tree was clean at
b2d6436(v1.5.1), in sync withorigin/main, CI green. No dirty state / merge conflict / uncommitted junk to recover. Proceeded with new work.
REA delta
- Pulled
morluto/rea:c0c390220→69bf61f7c(46 commits). - Headline new analyzer: #8154 external DEX method caller tracing
(
MAX_XREF_TEXT_BYTES= 8 MiB − 1 KiB,MAX_EXTERNAL_XREF_BYTES= that / 8). - Also noted: #8015 .NET field constant defaults (no new trip-able budget);
#8117 Apple
.appex/.framework/.xpc/.pluginbundle open (admission expansion, not a densify target tonight).
Done tonight
src/shields/formats/android-dex-xref.ts+ unsigned-APK wiring inarchive.ts.src/identity.ts+verify:rea(.java walk) + tests.- Release bla-agents@1.5.2.
Next step (2026-11-01)
- Empirical red-team: point sibling REA JADX external-references tool at a
BLA-hardened unsigned APK fixture; capture before/after in
docs/redteam/. - Optionally densify decoy callers further / add overload prototype bait for the method_inventory stage of #8154.
- Continue roadmap: ESP32 / Godot PCK / process-echo evidence deception, or MCP one-click polish.
2026-10-28 (Asia/Shanghai)
REA delta
- Pulled
morluto/rea:dd7a2da5b→0cd7ef456(325 commits). - New analyzers / budgets worth countering:
- #7456 caller-selected R8/ProGuard mapping (
PROGUARD_MAPPING_LIMITS: candidates=8, malformedSamples=16, projectionBytes=8MiB, indexBytes=128MiB;ANDROID_MAPPING_LIMITS.fileBytes=64MiB). - #7472 JNI native-declaration ↔
.soexport linking (ANDROID_JNI_REPORT_BYTES=16MiB,ANDROID_JNI_LIBRARY_LIMITS). - #7379 selected AndroidManifest component declarations (queued for next night).
- #7456 caller-selected R8/ProGuard mapping (
- Remainder of delta: evidence/contract refactors, browser/cli fixes — no new BLA-relevant budget constants beyond the Android set above.
Done tonight
src/identity.ts: registered Android mapping + JNI +MAX_NATIVE_EXPORT_ROWSlimits.src/shields/formats/android-r8.ts: mapping generator (residual conflicts + malformed bait) + ELF64 AArch64 ET_DYN decoy with realJava_*dynsym exports.- Wired into unsigned APK path + DEX trailer JNI names.
verify:reacatalog extended (Object.freeze-aware).- Release bla-agents@1.4.6.
Next step (2026-10-29)
- Empirical red-team: point sibling REA Android mapping + JNI tools at a
BLA-hardened unsigned APK fixture; capture before/after in
docs/redteam/. - Counter REA #7379 with binary/XML AndroidManifest component decoys (unsigned only).
- Optionally densify decoy JNI
.sopastANDROID_JNI_REPORT_BYTESprojection when intensity ≥ maximum (keep red-line: never rewrite developer libs).