Run the existing README consistency and Python producer guards from their
CI owners. Execute full local validation fresh while preserving cached
preparation, and permit scoped reuse of earlier portable build entries.
Collect failed-shard diagnostics before enforcing success and publish
compact, portable timing history only after a complete successful merge.
Preserve installer truncation failure across supported Bash versions.
Keep cancellation tests focused on the called MCP tool, observe server
progress before timing target readiness, and retain delayed-startup,
cancellation, cleanup, and live follow-up request assertions.
Validation: all 32 CI jobs and the separate Go lane passed; 935 files,
9,630 cases, 9,544 passed, 86 skipped, zero failed. Controlled regressions
and fault/restoration probes are recorded in PR #2210.
Refs #2190.
Assisted-by: Kogasa (gensokyo-skills:kogasa-surprise-testing)
* feat(flutter): identify Dart snapshot builds by pure APK parsing
Add a flutter-provider family with one operation, identify_flutter_build,
that reads an APK's lib/<abi>/libapp.so and libflutter.so through the
hardened ZIP reader and reports build facts as observations:
- The Dart VM snapshot hash, read 20 bytes after every snapshot magic
(f5f5dcdc). The vm and isolate sections are cross-checked: agreeing
sections collapse to one snapshot_hash, disagreeing sections keep every
candidate with snapshot_hash null and partial coverage.
- Per-ABI library digests and byte sizes, plus the engine library's GNU
build-id note and any labeled 'Dart VM version' or Android clang/LLVM
toolchain strings, verbatim; release builds without a version string
report null rather than a guess.
- The APK is digested before parsing so Evidence binds the observation to
the exact input, and a target with no libapp.so reports not-detected
with complete coverage.
The scanner is streaming with explicit pending-magic and rescan
bookkeeping, so chunk boundaries cannot double-count or drop sections, and
magic-free streams never accumulate. No engine, no external lookups: the
hash-to-SDK mapping stays with external datasets by design.
Verified against Flutter Gallery 2.9.2 (112 MB, three ABIs): every ABI
reports the same agreeing hash 65817c30a78bb44c3dc3771876b6010a from two
sections, per-ABI build-ids, and the clang toolchain line; a plain
aapt2-built APK reports not-detected with complete coverage. Scanner
goldens come from the same libraries. READMEs carry the new 'Flutter apps'
row in every supported language.
* feat(flutter): inspect Dart AOT snapshots through the ELF symbol table
Add inspect_dart_aot to the flutter family: for one ABI's libapp.so it
locates the four _kDart*Snapshot* symbols through the ELF dynamic symbol
table (ELF32 and ELF64 little-endian, bounds-checked reads; a payload
without exported symbols is refused), reports each section's offset and
size, validates data-section headers (magic, kind, length; instructions
sections honestly report nulls for the fields their layout does not
carry), and projects a bounded string pool from the isolate data section:
distinct package: source URIs (the app's dependency inventory), dart: SDK
URIs, and identifier-shaped tokens, each with counts, bounds, and partial
coverage when truncated.
The string pool is labeled as a byte-pattern projection, not deserialized
cluster semantics: typed class and function recovery remains an
SDK-specific parser, which the snapshot hash now unlocks. The ABI is
caller-selected with a deterministic default (arm64-v8a, x86_64, then
first available).
Verified against Flutter Gallery 2.9.2: all four symbols on both ELF64
(arm64-v8a) and ELF32 (armeabi-v7a) images, header validation, hash
agreement with identify_flutter_build, and a 756 package: / 201 dart:
URI pool from the isolate data section; goldens in DartElf.test.ts come
from the same libraries.
* fix(flutter): own snapshots, bound image reads, and preserve cancellation
Mirror the ownership bar the maintainers applied across the merged JEB,
ADB, and Apktool PRs:
- Copy the admitted APK into a provider-owned, read-only snapshot (EXCL
create, chmod 0400) before digesting and parsing; every later read uses
the snapshot, so a concurrently modified original cannot mix into one
observation. The result reports the snapshot's byte size and digest.
- Retain failed snapshot-root cleanup in a provider set and retry it on
close(), throwing an aggregate error when a root is still owned; root
removal is injectable so the retention is testable.
- Allocate the AOT image buffer at exactly the declared library size
before reading (bounded before allocation, not concat-after-stream),
keep the 256 MiB ceiling, and check cancellation between digest and
stream chunks through a captured abort reason.
- Accept APK paths containing spaces (covered by a fixture test) and
escape the ABI pattern's dash for v-mode regex compatibility.
Also add tests for cleanup retention with close retry and for
cancellation before any snapshot root is created.
* docs(flutter): cover the Italian README with the Flutter row
The Italian translation landed between the merged feature PRs; its table
now carries every merged row plus this one, keeping its documentation-link
set identical to README.md.
* fix(flutter): make the ABI pattern portable for v-mode validators
Use a hex-escaped hyphen in the ABI character class, matching the form
the ADB portability fix uses, so clients compiling advertised schemas in
unicodeSets mode accept the inspect_dart_aot input schema.
* refactor(flutter): reuse shared harnesses and deduplicate the vertical
The duplicate-detection gate (jscpd, 2.5% threshold) tipped to 2.51% with
the flutter files added. Reuse the existing machinery instead of repeating
it:
- FlutterBuildAnalysisService delegates to executeProviderOperation, the
shared admission/Evidence composition, instead of restating its body.
- The CLI registers both commands from one table-driven spec.
- Boundary tests use the shared serveCli and assertAdvertisedMcpContracts
harnesses rather than hand-rolled equivalents.
- The provider test consolidates the four payload-shape cases into one
table-driven case; the service test uses one admission table and
distinct phrasings.
Duplicate lines fall from 2.51% to 2.48%, back under the gate.
* fix(flutter): address review findings from real-app verification
CassieFox23 verified the PR against three real F-Droid Flutter apps and
a control app, and found five gaps. All are fixed and re-verified against
cone (info.tangential.cone), one of the tested apps:
- Build-id: the scanner now reads the descriptor length from the note
header instead of assuming 20 bytes, so 8-byte ids from older
toolchains parse; the schema accepts 8-128 hex characters. Verified
against cone's readelf-reported ids on all three ABIs.
- Compiler line: the toolchain capture bound rises from 200 to 500
characters, keeping cone's 262-character clang line verbatim.
- Zero-byte libapp.so: the extraction target is created before streaming,
so an empty entry yields a clean unsupported_target refusal for
inspect_dart_aot and elf_header_valid: false with partial coverage for
identify_flutter_build, instead of a raw ENOENT with code UNKNOWN.
- Truncated library: identify_flutter_build marks fewer than two snapshot
sections as partial coverage with a limitation naming the count, and a
missing ELF magic gets its own limitation.
- Docs: the default ABI preference now matches the code exactly
(arm64-v8a, x86_64, x86, armeabi-v7a, then the rest alphabetically).
Also: every test that read the author's /tmp/opencode goldens now builds
a self-contained synthetic ELF64 image with dynsym, four snapshot
symbols, two snapshot sections, and a string pool, so the suite runs
anywhere; the verify lane now requires a present hex build-id per ABI
instead of accepting null.
* fix(flutter): address the Codex review findings
Eight findings, all applied:
- P1: SnapshotHashScanner keeps candidates in a Set (no quadratic
Array.includes) with a 64-distinct-hash cap, so a hostile payload of
magic+hash records cannot stall the scan or bloat the response.
- P1: projectStringPool stops admitting new distinct entries once each
category reaches twice its projection bound and reports
collectionCapped; allocation and sorting now track the response, not
the payload.
- P2: ZIP reader rejections during entry iteration or entry streaming
map into the typed error algebra (cancellation vs input refusal)
instead of escaping as unhandled exceptions in both operations.
- P2: magic_valid now marks coverage only for the *Data symbols; a
complete four-symbol image with valid data magics reports complete
coverage (the instruction symbols intentionally lack the magic).
- P2: the MCP server closes the Flutter provider during shutdown, so a
retained snapshot root is retried and reported, not leaked.
- P2: LabeledStringScanner commits matches only when a printable run is
delimiter-terminated; the carry grows to 600 bytes so a 500-character
toolchain line survives any chunk boundary untruncated.
- P2: both Flutter tool effects declare writesFilesystem: true; the
owned APK snapshot and extracted libapp.so are temporary writes.
- P2: the verify lane asserts every ABI agrees on exactly one snapshot
hash instead of discarding the computed set.
* fix(flutter): address the second Codex review round
Nine findings on the first-round fixes:
- P1: identify_flutter_build enforces a cumulative extraction budget
(768 MiB) across every library entry, not just per library, so an APK
with many ABIs cannot stream an unbounded total.
- P1: a single NUL-free printable run stops materializing at 1024 bytes
and marks the pool capped, so an adversarial one-run section cannot
allocate unbounded strings.
- P1: ELF cString bounds the NUL-terminator scan to the name maximum
plus one, so malformed symbol tables cannot scan the whole image per
entry.
- P2: mapReaderFailure distinguishes ArtifactReaderFailure texts —
budget/limit rejections map to resource constraints, integrity
failures (CRC/overlap/signature) to their own input refusal — instead
of collapsing all reader failures into one message.
- P2: the labeled-string carry is cleared when a delimiter commits a
run, so consecutive toolchain lines are never concatenated.
- P2: string-pool counts are renamed *_count_at_least and documented as
lower bounds, because capped collection discards values beyond twice
the projection bound.
- P2: inspect_dart_aot marks hash disagreements, non-pair source counts,
and candidate-cap saturation as partial coverage.
- P2: only valid lowercase-hex hash records count as snapshot sources;
complete magic records with non-hex fields do not inflate agreement.
- P2: the per-entry size refusal in inspect_dart_aot moves after ABI
selection, so an oversized unselected ABI no longer refuses a small
request.
* fix(flutter): address the third Codex review round
Five findings:
- P2: the owned snapshot revalidates the source after the copy (size,
mtime, inode), so an in-place replacement or swap mid-copy refuses
the observation instead of publishing mixed bytes.
- P2: a failed ZipArtifactReader close retains the snapshot root for
close() retry instead of silently swallowing the descriptor leak,
in both operations.
- P2: snapshot headers are parsed within their symbol bounds and a
symbol smaller than the 16-byte header reports magic_valid false,
so trailing-image bytes can no longer fake a valid data section.
- P2: the verify lane counts provider.close() cleanup failures as
lane failures instead of certifying a run with retained roots.
- P2: LabeledStringScanner reports toolchain_capped when further
distinct matching lines are observed past the 4-line projection;
identify marks such results partial with a naming limitation.
* fix(flutter): address rounds three and four of the Codex review
Round three (one missed finding) plus round four (five findings):
- P2: the GNU build-id note descriptor size is decoded from all four
bytes of the little-endian field, rejecting nonzero high bytes
instead of masking them to 16 bits.
- P1: the Flutter MCP handlers route through createContractToolHandler
and the session's analysisAdmission, so close_binary and snapshot
operations drain in-flight Flutter calls instead of racing them.
- P2: the owned-snapshot copy is a streamed, abort-checked write (no
copyFile), so cancellation stops the write promptly while the
source-identity revalidation (size, mtime, inode) still guards
against mid-copy replacement.
- P2: identify marks any snapshot_hash_sources value other than exactly
two as partial, matching the inspection path's pair requirement.
- P2: the verify lane refuses to run without --apk (exit 2) instead of
printing a false-green pass, and the engine check requires every
libapp-carrying ABI to have a present libflutter with a valid hex
build-id rather than filtering missing engines away vacuously.
* fix(flutter): address the fifth Codex review round
Six findings plus the upstream duplication baseline:
- P2: the flutter-provider kind reports available in the capability
inventory, so binary_session no longer lists always-registered
Flutter tools as provider_missing.
- P2: printable runs that exceed the 1024-byte cap are discarded at
their delimiter instead of emitting their truncated prefix as a
fabricated observation.
- P2: ELF symbols must fit inside the file-backed portion of their
PT_LOAD segment (p_filesz), so addresses in gaps or BSS are refused
instead of accepting unrelated bytes.
- P2: the build-id carry spans the complete 80-byte maximum note
(16-byte header + 64-byte descriptor) so chunk-boundary splits
cannot drop the first byte.
- P2: the labeled-string carry is sized for the toolchain pattern's
true 605-byte maximum, not the previous 600.
- P2: the AOT completeness predicate requires exactly two sources and
one candidate unconditionally, so zero-source images with valid data
magics report partial instead of complete.
Also: the shared ELF fixture builder and snapshot-section helper moved
to a dedicated FlutterTestFixtures module, eliminating cross-file test
duplication; jscpd now passes at 2.50% while main sits at 2.52%.
* fix(flutter): address the sixth Codex review round
Five findings:
- P2: snapshot-hash scanning is bounded to the exported data symbols'
byte ranges, so decoy or stale magic/hash sequences elsewhere in the
image cannot inflate source counts or contribute candidates.
- P2: labeled-string runs longer than the carry bound are rejected
instead of truncated, preventing a regex prefix from being published
as a verbatim source string.
- P2: the AOT image reread checks cancellation between file reads and
before the synchronous parsing stage, so a cancelled caller cannot
receive a success from a completed 256 MiB read.
- P2: the verify lane requires exactly two snapshot sources per ABI,
matching the provider's pair requirement, instead of accepting any
count >= 2.
- P2: ELF section and program header entry sizes are validated against
the class-defined minimum widths before walking either table, so
undersized entries cannot cause field reads to cross into
neighboring bytes.
* fix(flutter): address the seventh Codex review round
Six findings:
- P2: data-symbol ranges are deduplicated by offset and size, so aliased
symbols mapping to the same bytes are scanned once and the result
marks partial coverage with a limitation naming the shared range.
- P2: openEntrySafely forwards return/throw to the underlying Readable's
iterator, so an early for-await exit closes the ZIP entry stream and
releases its producer instead of leaking a backpressured pipe.
- P2: inspect_dart_aot now reports snapshot_hash_candidates and
snapshot_hash_sources alongside snapshot_hash, so a null hash from
conflicting evidence is distinguishable from missing hash data.
- P2: overlong printable runs only set toolchain_capped when they begin
with a recognized label prefix (Dart VM version: or Android (); other
long strings such as license text are skipped without marking partial.
- P2: readElfSymbol falls back to PT_DYNAMIC/DT_SYMTAB when section
headers are absent (e_shnum 0, ET_DYN), so stripped Flutter libraries
with exported snapshot symbols are no longer refused; the dynamic
table is bounded by DT_STRSZ, DT_SYMENT, and the image end.
- P2: the verify lane exercises inspect_dart_aot for every ABI carrying
libapp.so, not just the preferred default.
* fix(flutter): address the eighth Codex review round
Five findings:
- P2: each data-symbol range gets a fresh SnapshotHashScanner so the
carry from one noncontiguous slice cannot fabricate a record spanning
into the next; results are aggregated after scanning.
- P2: dynsym name lookups are bounded to the linked string table —
st_name offsets outside the table and terminators beyond it are
rejected instead of reading unrelated payload bytes as symbol names,
in both the section-header and PT_DYNAMIC paths.
- P2: identify_flutter_build routes its reader.entries and reader.open
calls through the same collectEntriesSafely and openEntrySafely
wrappers as inspect_dart_aot, so malformed ZIPs surface as input
refusals instead of AnalysisProtocolError.
- P2: mapReaderFailure recognizes ArtifactReaderFailure texts carrying
a cancelled reason, mapping them to AnalysisCancelledError instead
of AnalysisInputError.
- P2: the cumulative extraction cap is enforced inside the chunk loop,
not after each library completes, so refusal happens before the
remaining bytes are streamed and hashed.
* fix(flutter): harden streaming, sizes, and cancellation after self-review
Four issues found in a systematic self-audit across all bug classes
from eight Codex rounds:
- createReadStream in both the digest function and the owned-snapshot
copy now receives the AbortSignal directly, so cancellation aborts
the filesystem read at the source instead of waiting for the next
chunk to arrive on a slow filesystem.
- The owned snapshot enforces a 2 GiB APK-level size cap before any
copy I/O, so a path pointing at an enormous file is refused before
disk usage grows.
- identify_flutter_build preflights the sum of declared entry sizes
against the 768 MiB cumulative budget, refusing before any ZIP
extraction when the declared total already exceeds it.
- The LabeledStringScanner triple result() call is collapsed to a
single invocation (the previous code called result() three times in
the emission, relying on end() being idempotent).
No behavior change for valid inputs; all 36 tests, check:fast, and
the cone verify lane pass unchanged.
* fix(flutter): require structurally parseable ELF headers for complete coverage
CassieFox23's third corrupt-header case: an ELF with valid magic but
zeroed section/program header offsets passes the 4-byte magic check
and the byte scanner still finds hash patterns in the data sections,
so identify_flutter_build reports complete coverage for a structurally
broken file.
The elf_header_valid check now also verifies beyond the magic:
e_type must be a known value (ET_REL/ET_EXEC/ET_DYN), at least one
of e_shoff/e_phoff must be non-zero, and at least one of
e_shnum/e_phnum must be non-zero. A file with valid magic but zeroed
structure reports partial coverage with the existing ELF-magic
limitation.
* fix(flutter): validate ELF and snapshot coverage
Share bounded ELF header parsing between identification and inspection. Accumulate headers across ZIP chunks and validate table entry widths and extents before claiming complete identification coverage.
Bind AOT hashes to the exported VM and isolate headers, validate stored snapshot lengths including the magic bytes, and preserve partial evidence for missing, malformed, aliased, or overlapping data headers.
Preserve native stream cancellations and APK resource constraints across both operations. Bound snapshot copies while streaming and retain useful input-failure diagnostics and lower-bound count wording.
Validation: 72 focused tests; check:fast; check including formatting; docs:check; test discovery; jscpd; real cone APK and non-Flutter control on Linux. Verified CLI/MCP parity for valid and corrupted ELF headers and AOT inspection across three ABIs. Native Windows and macOS were not run.
* fix(flutter): unify artifact ownership and bound ELF evidence
* fix(flutter): preserve ELF provenance and partial coverage
Read GNU build IDs only from bounded ELF note ranges and retain conflicting candidates. Preserve source versus owned storage failures and honor cancellation through parsing and final cleanup.
Separate snapshot structure from string-pool coverage. Verify every Gallery ABI with digest-bound engine labels and URI counts, and document the supported snapshot-kind layout.
---------
Co-authored-by: REA Contributor <contributor@rea.tools>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Put setup, developer commands, MCP replies, and investigation recipes before
client details and implementation references. Preserve existing guide anchors,
examples, provenance, partial-result recovery, and provider restrictions.
Correct verified documentation drift, synchronize the Ghidra capability
wording, and regenerate metadata for skill version 40. Keep uncertain platform
and lifecycle findings unchanged. Refs #2156.
Validation: 56 focused tests, check:fast, docs:check, VitePress production build,
format and link/example checks. Real isolated CLI/MCP probes cover Ghidra,
NativeAOT, HAR, cancellation, Chrome, large ELF delivery, native type limits,
and the new JS/managed/process/setup recipes. Heavy work uses two CPUs;
a resource-heavy DOM probe was stopped and replaced by a smaller real CDP
payload-limit control. Windows, macOS GUI, ADB, and the full CI matrix were
not executed.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
* ci: narrow native-host lane selection and reuse build caches
Unowned src/ and bridge/ files no longer fall back to the full baseline.
They run the Linux suite plus the Linux x64 installed-package/Inspector
baseline. New apple and package scopes own the Apple artifact lane, the
package matrix and the Windows curated lane, including the test files those
lanes invoke. Manual-only verifier entrypoints run static checks and the
Linux suite. CHANGELOG, SECURITY, GitHub templates and the native Windows
README are documentation-only.
Pull requests omit the macOS x64 package host unless labelled ci:full.
The Windows and Apple lanes restore the shared Turbo build cache, the
Windows lane drops its duplicate typecheck, and the Linux suite uses six
shards. Main pushes keep the full baseline.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* ci: match package script entrypoints against scope globs
Script-only package.json edits resolved their entrypoint with literal
equality, so glob-owned verifiers such as scripts/verify/apple/** and
scripts/verify/inspector/** fell back to the full baseline instead of
selecting their owning lane.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat(setup): register DeepSeek Harness through setup, doctor and uninstall
DeepSeek Harness registers MCP clients through a Cordis YAML patch
sequence in the home patch rather than a server table, so setup could
not configure it and users had to append the documented entry by hand.
Add a dsh client: the home patch $DSH_HOME/cordis.patch.yml
(~/.dsh/cordis.patch.yml when unset) is parsed as a patch list, the REA
entry owns the stable rea-mcp id, and setup appends or replaces exactly
that entry while preserving unrelated patches and comments in place.
Doctor reports the registration through the same entry, and uninstall
removes it, dropping insert operations it emptied. Skill destinations
follow ~/.agents/skills ($DSH_AGENTS_HOME/skills when configured), the
home the rest of the client machinery already uses.
Fixes#1839
* fix(setup): carry environment and identity validation into DSH registrations
Address review: the DSH entry discarded discovered provider overrides;
config.env is now emitted and retained like the other dialects. An entry
carrying the rea-mcp id under a wrong or missing package name, or
duplicated entries, reads back as absent so setup rewrites exactly one
canonical entry instead of reporting unchanged. Doctor now requires
serverName rea and stdio transport before calling a DSH registration
aligned, and the Harness skill identifier in the docs matches the
installed reverse-engineer-anything skill.
* fix(setup): apply Cordis patch order and collapse DSH registrations
Address review: a later direct { id, config } operation replaces the
row's config under the native patch DSL, so the registration view now
applies patches in order instead of only reading insert operations;
doctor reports the effective command rather than the inserted one.
Setup rewrites exactly one canonical entry: duplicate id-matching rows
collapse to the first, direct patches to the entry are dropped from the
document and from the edited YAML, and uninstall removes them with the
entry. The client-location test fixture now clears DSH_HOME and
DSH_AGENTS_HOME so a developer environment cannot point the suite at
real Harness homes.
* fix(setup): repair direct-only registrations and preserve Cordis entry state
Serialize now edits the first insertion carrying rea-mcp in place, so a
grouped entry keeps its scope, neighbors, and comments instead of moving
to a fresh root-level insertion, and later carrying insertions and direct
patches collapse away. Entry-level disabled state projects into the
registration the doctor and setup compare, so a disabled row reads as
stale and the repair re-enables it. A direct-only registration no longer
suppresses the canonical insert append, and an update retains the
registration environment. Duplicate entries collapse across operations,
not only within one.
* fix(setup): preserve Harness row state through identity and refresh repairs
A wrong-package row still reads back as a server record carrying its
environment and caller fields, projected with a package marker the
alignment path treats as stale, instead of vanishing and dropping
config.env from the repair. A direct Cordis patch without a preceding
insertion configures nothing, so a direct-only registration reads as
absent even when its config matches and the repair appends the insert.
Caller-owned config fields beyond the REA-owned keys, such as the cwd
the native verifier sets, survive a refresh instead of being silently
dropped.
* fix(setup): preserve Cordis tags and collapse aligned Harness rows
A refresh rebuilt Harness config from parsed values, so a caller !!js
expression was written back as a literal. Owned keys are now overlaid on
the winning config node and other nodes stay in place. Two already-aligned
rea-mcp insertions read as stale, so setup collapses them to one row, and
the bundled skill lists dsh with the other clients.
* fix(setup): drop the unused DSH configuration type
Knip reports DshConfigurationFormat as an unused export. The format union
does not depend on that alias.
* fix(setup): keep Harness env tags and comment-only patches
A refresh replaced the whole env map, so a caller !!js value was written
back as a literal. Environment entries now stay in place unless the value
changed. A comment-only cordis.patch.yml parses as an empty patch list and
keeps those comments when the registration is appended. Docs describe
plain rea uninstall, which scans supported clients and has no --client flag.
* fix(setup): own Harness rows by entry id and expression state
Uninstall removes a rea-mcp row even when the launcher is stale or the
patch never established a server record. A non-boolean disabled value,
including a Cordis expression, reads as stale. Expression text on a known
environment setting is not validated as a literal path, so a refresh can
keep the tag.
* fix(setup): collapse aliased Harness insertions
An anchored insert followed by an alias parses as two rea-mcp rows, but
the YAML edit kept the alias, so readback stayed duplicated and setup
restored the file. Aliases that resolve to the owned entry are now
dropped, including an alias of the entry inside one insert. Aliases of
other patches stay.
* fix(setup): resolve aliased Harness lists and config maps
An insert value that aliases another entry list still loaded a second
rea-mcp row after the operation-level alias collapse. That alias is now
dropped once the anchored list is rewritten. A config: *cfg map is
resolved and cloned before the refresh, so a caller !!js field on the
aliased map is not rewritten as a plain literal.
* fix(setup): follow Harness aliases for ids, lists, and env tags
An insert alias whose anchor is not another insert is copied so
unrelated rows stay while the owned row is rewritten. An aliased
entry id is resolved before ownership checks, and an environment alias
of an unchanged !!js scalar is left in place.
* fix(setup): keep aliased Harness env maps through uninstall
An env alias is cloned before a refresh so tagged values are not rebuilt
as literals, and expression-key scans follow aliases before path checks.
Insert aliases that carry REA are recorded before uninstall empties their
anchor, so the later alias is dropped instead of dangling.
* fix(setup): follow root Harness aliases and mark dsh as main-only
Winning-config walks resolve a root alias before patch order, so a replayed direct config's !!js env is not validated as a literal path. Root aliases that own rea-mcp are recorded before uninstall empties their anchor. An insert whose only owned member is an external alias is cloned so Cordis tags survive. Docs state dsh is not in npm 6.4.0.
* fix(setup): keep rows behind inert Harness aliases
A root alias whose anchor is not a loaded operation is copied and only the owned row is edited, so unrelated rows stay active. Entry id aliases become a scalar before their anchor patch is dropped. Root-alias ownership follows insert-list aliases, so a replay is removed instead of dangling after uninstall.
* refactor(setup): centralize Cordis source ownership
* fix(javascript): share Electron mutation proof with wrapped calls
* test(mcp): align graph and catalog checks with current contracts
* fix(setup): separate Cordis ownership from editability
* docs: clarify ownership and interpretation boundaries
* fix(setup): preserve Cordis applicability and caller metadata
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* test(clients): verify native Gemini CLI integration on Windows
The Gemini CLI client lane asserted POSIX. On Windows it also needs the npm install's JavaScript entry point, because only a gemini.cmd shim reaches PATH, and two setup-plan checks compared paths with a hard-coded slash. Remove the assertion, run a selected .js, .mjs or .cjs entry point under the verifier's Node, build those prefixes with path.sep, name the entry point in the Windows preflight error, and record 0.63.0 as verified on native Windows in docs/testing.md.
Fixes#1935
* test(clients): resolve the selected Gemini CLI entry point once
The preflight runs from the caller's directory and the client run from the disposable workspace, so a relative REA_VERIFY_GEMINI_COMMAND passed the preflight and then failed with Cannot find module. Resolve a selected JavaScript entry point to an absolute path before both launches.
* fix(wasm): preserve debug module names in WABT output
* fix(native): accept otool dylib-use options
Preserve modern dylib options while requiring one encoding metadata row and retaining duplicate and multiline-output checks. Share field parsing between preservation and validation.
Verify delayed dylibs through real macOS CLI/MCP workflows and refresh the high-address golden with the provider's -h -l capture. Validated 85 focused native tests, npm run check, and verify:macos-bundle on macOS arm64.
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* fix(wasm): accept unpadded WABT section rows
WABT 1.0.42 right-aligns section names in a nine-column field. DataCount
fills that field, so its objdump row has no leading whitespace. Accept
that legal row. Header recognition, the required field shape, and section
range checks still reject malformed output.
Fixes#1956
* fix(ci): refresh MIPS expectations and Windows test paths
Assert admitted soft-float ABI metadata and its committed storage policy through the public target resolver, retain FP64 rejection, and point the curated Windows lane at the moved filesystem reconstruction tests.
Validation: reproduced all four prior failures; 97 focused tests and check:ci pass. Native Windows and real Ghidra execution were not performed locally.
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Apply caller-selected analyzer opt-outs before Ghidra analysis, reject unknown or ineffective names, and commit raw/header imports to the same analyzer profile. Use supported x86 raw compiler specs and cover the real analyzer/readback and raw workflows.
Adopts the original implementation from #1800 with validated corrections. Refs #1783, #1836.
Co-authored-by: akushonkamen <74654766+akushonkamen@users.noreply.github.com>
Co-authored-by: N0zoM1z0 <161784452+N0zoM1z0@users.noreply.github.com>
Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com>
The Claude Code client lane asserted POSIX, although nothing in it depends
on it. Remove the assertion, tell Windows users to select claude.exe when
the default command cannot be spawned (npm installs put only a claude.cmd
shim on PATH), and record 2.1.296 as verified on native Windows in
docs/testing.md.
Fixes#1923
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* feat(native): admit raw binaries with caller-supplied architecture and entry
* fix(native): honor raw-binary preparation and entry identity
* fix(ghidra): stream raw snapshot attestation with cancellation
* fix(native): preserve and validate raw lifecycle entries
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Share CLI boundary module evaluation in bounded forks while preserving fresh command instances and child processes. Keep module-mocked cleanup tests and source/compiled product catalog verification isolated, and update focused artifact selection.
Flameox full cohort with V8 coverage: 102.8s to 96.6s; 290 passed and two existing skips. Statement, function and branch coverage maps and hit sets match across all 1,153 source files. After syncing main, 75 focused tests, typecheck, lint, docs checks and formatting pass; all 834 test files are discovered exactly once.
* feat(harmony): project HarmonyOS package inventory evidence
First HarmonyOS increment from #1431: HAP, HSP, and App Pack roots are
classified through the hardened ZIP reader and projected into an
execution-free application graph with packaging-model hints, exact
component paths and hashes, App Pack children by path, runtime-family
hints, and path-based N-API bridge hypotheses. HarmonyOS .har stays
unclassified by suffix because .har also names HTTP Archive JSON.
Adds the project_harmony_application_graph tool (CLI and MCP), synthetic
composition coverage, a pinned real-package verify lane, and docs.
* refactor(domain): share application graph projection scaffolding
The HarmonyOS projection mirrors the Android and Apple projections by
design, which pushed jscpd duplication over its 2.5 percent budget.
Extract the identical pieces (inventory input schema, projected
component schema and occurrence flattening, bridge candidate schema
factory, coverage assembly, comparator, canonical digest) into
domain/mobileApplicationGraphProjection.ts and route all three
projections plus the composition fixtures through it. Duplication drops
from 2.52 to 2.49 percent with no behavior change: projection ids,
classifications, and limitations are byte-identical.
* fix(harmony): address review, stable app-pack roots, and rotated fixture
- rebased onto main; the shared projection scaffolding now keeps main's
nativeRuntimeConvention loop and the harmony runtime families adopt the
same react-native/flutter/unity path conventions
- the .app App Pack suffix is now a root-input rule only (verified ZIP
magic in classifyRootFormat and BinaryTargetResolver); nested .app ZIP
members inside other packages keep the generic zip classification, with
a regression test, so existing inventories do not change
- the signing limitation now states that an empty signing list does not
mean unsigned because packaging-tool signatures live in the ZIP signing
block, and the app-pack root rule is disclosed as suffix-based
- repinned the real-package fixture to the current VHome 2.6.14-beta asset
after the upstream deleted v5.0.111-beta and replaced the v2.6.14-beta
asset within hours; the pin records both observed digests
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* fix(bun): report the real runtime and enforce worker memory budgets on Windows
Under Bun, REA recorded Node/V8 identity it did not observe and its JavaScript helper processes ran without a memory bound, because Bun ignores the V8 heap flags REA passes. Source-map trace Evidence claimed runtime node-v8 with Bun's synthetic V8 heap limit (the decoder then rejected that limit, so trace_web_source_location always failed), analysis resource failures reported heap limits Bun does not enforce, the HAR decoder claimed a fixed 192 MiB heap, and doctor reported Bun's Node.js compatibility version as the Node version.
Identify the runtime once from process.versions. Source-map reports record id bun with the Bun and Node.js compatibility versions and a null v8_heap_limit_bytes; the decoder checks the heap budget only for V8 reports and rejects a report from a runtime other than the launching one. The analysis worker reports a null heap limit under Bun, HAR limitations and failures omit heap sizes under Bun, and the doctor node check reports Bun and passes when its Node.js compatibility version is supported.
On Windows under Bun, each helper's owned Job Object now carries a commit limit of its heap budget plus a measured 512 MiB Bun runtime reserve (704 MiB for the codec and HAR decoder, 1536 MiB for the default analyzer). The native addon accepts an optional page-aligned job memory limit on process_spawn, verifies it by readback, associates a private completion port before the child is created, and reports a sticky memoryLimitReached and peak job commit on each poll; inspect reports jobMemoryLimit. A requested limit with an artifact lacking the control refuses the launch, and non-Windows launchers refuse a limit.
Reaching the limit is classified by the job notification rather than Bun's exit status: the analyzer fails the current source with a memory resource failure reporting worker_process_memory_limit_bytes, the codec returns a memory resource failure, and the HAR decoder returns its memory failure with process_memory_limit_bytes. Bun source-map Evidence records process_memory_limit_bytes, verified against the launcher. Node.js launches and outputs are unchanged.
* fix(windows): drain memory notifications after observing exit
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(javascript): recover code-split bundles from multiple inputs
recover_javascript_sources accepted a single path, and
validateModuleSet bound every module's provenance to that one snapshot.
A code-split app (runtime + chunks, or an Electron main/preload/renderer
set) had to be recovered one file per call, and modules recovered alone
could not resolve imports across chunks: cross-chunk require(id)
references stayed as numeric ids with property reads.
Add an optional additional_paths input. The listed files are
concatenated, in order after path, into the single engine snapshot, so
one Wakaru run sees every module and cross-chunk imports resolve.
validateModuleSet's single-snapshot binding is unchanged. Module
provenance is attributed back to the input file its first extraction
range starts in, with byte ranges rebased to that file's own offsets;
a module whose ranges cross an input boundary keeps the concatenated
snapshot identity and offsets, recorded as an explicit limitation.
Single-path behavior is byte-for-byte unchanged (one segment,
offset 0).
Fixes#1605
* fix(javascript): preserve source identity in multi-input recovery
---------
Co-authored-by: JingHao-Leon <w15909130000@gmail.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Transfer the build job Turbo cache to the four native package runners within the same workflow. Keep the normal build, prepack, pack, fresh installation, updater, and Inspector verification commands. Native artifacts and dependencies remain host-local.
Validation: actionlint 1.7.7, check:ci and docs:check passed. A fresh independent checkout restored 1124 byte-identical files from 1.64 MB of compile/skills cache in 163 ms; a source change invalidated compile and skills entries. Full package and real Node Inspector verifiers passed on macOS arm64. Cross-platform CI timing remains to be measured.
* test(ida): expect a transient busy worker health to be waited out
* fix(ida): wait out a transient busy worker health before reading identity
* docs(ida): record the verified macOS arm64 headless workflow and the busy-health wait
* test(ida): pin the busy wait deadline, cause and reported tool, and assert clean release
* fix(ida): repeat only a tool-name-shaped busy tool and keep the wait in its own module
* test(ida): drive the busy wait with a virtual clock and pin its bound and cancellation
* fix(ida): stop the busy wait at its deadline without another read and honour a late abort
* docs(ida): state when the busy wait times out, cancels or reports an unknown tool
Run the 22 malformed process scenarios in pairs instead of starting 22 compiled CLI runtimes simultaneously. Retain every CLI/MCP error assertion and the no-launch check, and document nested subprocess limits.
* feat(ghidra): recover signed AArch64 relative jump tables
Generalise the AArch64 relative-table matcher to signed byte/halfword entries, UXTW indexes, byte or LSL #2 offsets, B.CS/B.HS bounds and a CBZ route for zero entries. Cross-check it against Ghidra's own load table when one exists, and add a fixture modelled on the RenderScript blend kernel.
Refs #487
* fix(ghidra): preserve typed labels for normalized switch selectors
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(ghidra): configurable import language and analysis seeds
REA_GHIDRA_LANGUAGE_ID and REA_GHIDRA_COMPILER_SPEC_ID set the import
language when the header cannot identify the compiler (for example,
Borland C++ PE images).
REA_GHIDRA_SEED_FILE seeds functions, code and labels before
auto-analysis, for code reached only through computed jumps or data.
Both are validated before launch, recorded in the analysis profile and
verified in the bridge handshake.
* fix(ghidra): parse language XML and preserve seed ordering
* test(ghidra): use explicit platform paths for seed launch checks
* fix(javascript): retain cleanup failure reason in error causes
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Use password-store=basic for owned raw Chrome verification processes and document the same option for dedicated caller-owned Linux investigation profiles. Locked keyrings can stall navigation and leave placeholder frame URLs; keep the existing committed-URL authorization boundary.
Validation: real Chrome 154 CLI and stdio MCP browser verifier passed on macOS ARM64, plus docs freshness and check:fast. The locked Linux desktop keyring environment was not reproduced locally.
Bind HOME and USERPROFILE in disposable native client verifiers, exercise default personal skill discovery, and clarify automatic Pi/Hermes setup availability in the checked npm release.
Validation: integrated build/static/docs checks, repeated native client workflows, real Codex-backed client trials and two independent cold-npx Codex installs. All 19 CI checks passed on a7ba07ba. Other model providers, platforms and deep-analysis engines retain their explicit verification TODOs.
Guide task-centered recovery from observed client, registration, PATH and
provider blockers. Install the recovery reference, document exact JavaScript
module selectors, and require clarification before searching for an unnamed
app. Validate public setup in isolated Codex accounts and preserve failed
or unrecognized tool outcomes in the existing evaluator.
Refs: #1587
BREAKING CHANGE: Optional verify:agent reports now use schemaVersion 4.
Consumers must accept expectedFirstTool: null and targetClarificationPassed
for missing-target scenarios. Production CLI/MCP contracts are unchanged.
* feat(apktool): decode Android resource tables through a caller-supplied launcher
Add an Apktool family with two operations over the caller's apktool
launcher (REA_APKTOOL_COMMAND, absolute, or apktool on PATH):
- inspect_apktool_client: resolved launcher, selection source, and
version; decodes nothing.
- decode_android_resources: apktool d --no-src in a provider-owned
workspace that is removed before the result returns. Projects the
decoded AndroidManifest.xml (bounded at 512 KiB), apktool.yml version
and SDK facts exactly as printed, string resources verbatim (default or
one locale, bounded at 2000 entries, nested-markup entries counted as
unparsed), available string locales, and the decoded file count with a
hard cap. The target APK is digested with SHA-256 before decoding so
Evidence binds the observation to the exact input.
Smali production is skipped by design: Java-source recovery belongs to
the JADX family. Workspace cleanup happens before results return; an
unremovable workspace is named in the limitations instead of abandoned.
Verified against apktool 2.7.0-dirty (Debian packaging) on Linux with
OpenJDK 25 against a signed aapt2-built probe APK with two resource
locales: launcher version, on-disk digest agreement, metadata projection,
manifest package agreement, and both string tables (default and de).
Parser goldens come from the same decode. READMEs carry the new
'Android resources' row in every supported language.
* fix(apktool): own snapshots and bound decoded resource reads
* docs(apktool): describe owned resources and bounded projection
---------
Co-authored-by: REA Contributor <contributor@rea.tools>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(jeb): add JEB engine provider over its MCP interface
Add a bring-your-own JEB provider that serves project, unit, and
decompilation inspection from a caller-started JEB client speaking MCP
(streamable HTTP, selected with REA_JEB_MCP_URL). Expose four provider-neutral
operations through shared CLI/MCP workflows: inspect_jeb_client,
open_jeb_project, list_jeb_units, and decompile_jeb_item, each returning
Evidence with engine identity, endpoint, and preserved refusal reasons.
The adapter keeps JEB's envelope semantics: success fields are parsed
format-aware at the boundary, message-only refusals become unsupported-target
errors with the engine's own reason, transport failures become availability
failures with recovery guidance, and unit pagination derives coverage from the
engine's 100-item page cap. Explicit unknowns (implicitly selected unit,
omitted filters) are reported instead of guessed.
Verified against a real JEB 5.48.0 licensed build serving jeb-mcp-server 1.3.0
headless: client inspection, unit listing, and method decompilation were
exercised through the CLI, MCP boundary, and scripts/verify-jeb.mjs.
JEB 5.48.0 headless instances do not advertise open_project; the refusal is
preserved and documented.
* fix(jeb): verify against a real signed APK and fix CLI option surface
Verify the provider against a locally built, v1/v2/v3-signed probe APK
(binary manifest, classes.dex with a launcher activity) on JEB 5.48.0: the
full unit tree (apk, manifest, three certificate units, dex bytecode,
decompiler), wildcard-filtered and paginated unit listing with derived
coverage, parent-restricted listing, and launcher-activity method
decompilation through both CLI and MCP, including dexdec's inline
constant folding of the probe's obfuscated tag.
Move list-jeb-units filter/pagination and decompile-jeb-item unit selection
from positional arguments to options; the MCP schemas are unchanged.
* fix(jeb): preserve engine refusals through the real MCP boundary
Testing against the published Signal 8.30.3 universal APK exposed a boundary
bug the scripted provider tests missed: decodeJebToolResult raised a protocol
error for isError results before the provider could map JEB's
{success:false, message} envelope, so real engine refusals surfaced as
unreadable output instead of the engine's own reason. Decode refusal
envelopes like any structured result and let the provider preserve the
reason; cover the decoder directly, including contentless and unparseable
results.
Signal verification: full 109 MB R8/Kotlin/Compose APK with four signature
schemes and native arm64 ELF units parsed; MainActivity.onCreate and R8
synthetic lambdas decompiled through CLI and MCP; paginated, filtered, and
parent-restricted unit listing; refusal path confirmed on the live engine.
* docs(jeb): record Flutter Gallery verification on the real-engine lane
Verify the provider against the official Flutter Gallery 2.9.2 release APK
(112 MB): per-ABI libapp.so Dart AOT snapshot images analyzed as native code
with 75k methods on arm64, snapshot internals and wrapper methods decompiled
through both CLI and MCP with explicit unit selection, decoded manifest, and
filtered native-library unit discovery. Note that JEB 5.48.0 headless does
not expose the dedicated Dart snapshot processor or run_script; both remain
GUI-only engine surfaces.
* docs(jeb): translate the README JEB row for every supported language
The JEB row was inserted in English across all translated READMEs; each now
carries its own label, capability summary, requirements, and guide link
following that language's Android-row conventions.
* fix(jeb): enforce local endpoints and preserve cancellation
* fix(jeb): refuse redirects across the local MCP boundary
---------
Co-authored-by: REA Contributor <contributor@rea.tools>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* fix(javascript): accept the Wakaru 1.x release line and verify 1.14.0
Exact --version equality rejected every Wakaru build except 1.13.0, and
strict report schemas rejected any added field, so each upstream release
broke recovery until REA re-pinned. Wakaru now documents that --version,
--json and provenance.json change only by addition within a major version.
Admit ^1.13.0 with the verified/compatible split used for Binwalk and
Unblob, report the observed version instead of the pinned constant, and
add a limitation for unverified releases. Parse reports with loose
objects and keep the reported module status as a string; path, identity,
range and file-set checks stay exact.
Move the audited pin, submodule, CI binary and verifier to 1.14.0, and
document where to get the binary from the refusal message.
* fix(javascript): project known Wakaru warning fields
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(adb): acquire installed APK sets from connected Android devices
Add a read-only ADB device-acquisition family over the caller's adb binary:
inspect_adb_client, list_adb_devices, inspect_adb_device, list_adb_packages,
and pull_adb_package, exposed through the CLI and MCP with shared admission,
Evidence provenance (binary, version, serial, transport), and fixed argv
vectors that never reach a shell.
pull_adb_package acquires the complete split set reported by pm path into a
package-named directory that must not exist, digesting every file with
SHA-256 and keeping per-file failure reasons and partial coverage. Device
kind is reported as an inference with its basis; getprop whitelisting keeps
build identity bounded. REA installs nothing and never manages devices or
emulators; the first device query may start the caller's local adb server,
recorded when observed.
Verified against adb 34.0.5 on Linux with an Android 14 (API 34) x86_64
emulator: full 224-package inventory, a real two-APK split set pulled with
byte-exact digests, and a 73.9 MB system-package pull. npm run verify:adb
records the lane; parser goldens come from the same device. READMEs carry
the new row in every supported language.
* feat(adb): complete the adb surface with observation, transfer, and lifecycle tools
Extend the ADB family from five to twenty-four operations, all through
fixed adb argument vectors with no shell exposure:
Observation: read_adb_logcat (bounded dump, count/buffer/pid), inspect_adb_package
(dumpsys package projection), list_adb_processes (ps -A), list_adb_directory
(ls -la with kinds and symlink targets), list_adb_features (hex GL versions),
list_adb_services (AIDL slash-suffixed names), inspect_adb_display (wm size and
density), inspect_adb_window (focus declarations), read_adb_setting (device-side
nulls reported as such), capture_adb_screen (exec-out screencap with PNG
dimensions), collect_adb_bugreport (digested device archive).
Transfer and search: pull_adb_file and push_adb_file (device-side sha256sum
verification, explicit overwrite authority), resolve_adb_packages (substring
package search with exact-match flag — the find step of an easy get).
Lifecycle for dynamic analysis, each declaring mutatesTarget:
install_adb_package and uninstall_adb_package preserving the device's
Success/Failure verdict, start_adb_app (resolve-activity then am start -W),
start_adb_activity (typed extras as --es/--ez/--ei/--el/--ef), and stop_adb_app
(am force-stop).
Verified end-to-end against adb 34.0.5 on Linux with an Android 14 (API 34)
x86_64 emulator: 287 services, 92 features, split-set pulls with byte-exact
digests, verified push/pull roundtrip, sized screen capture, and the full
install → resolve → start → observe-in-ps → force-stop → uninstall lifecycle.
Fixed a v-mode regex incompatibility in the settings key pattern and added
argument descriptions required by the CLI catalog audit.
* fix(adb): quote remote shell operands and protect captured output
* test(adb): cover bugreport destinations containing spaces
---------
Co-authored-by: REA Contributor <contributor@rea.tools>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Ysopwnable
Quentin Limo
Daniel Gonçalves Araujo
Kaoru0822
BaaRai
pablodiazjorge
Arthas
Nithilan Kumaran
morluto
Matt Van Horn
Thang Luu
xeropresence
JingHao-Leon
Mykola Dementiev
Syed Rishad
sqmagellan
tyloryang
Pionxzh