Use exact discovered test identities, production MCP sessions and byte-exact setup backups. Add CLI signal cleanup coverage and independent Ghidra, Go and firmware verification assertions; prune redundant test configuration and preserve platform limits.
Share tagged readiness observations, retain partial provider facts, authenticate bridge identity, bound diagnostics, and preserve cancellation and lost-session reasons. Fail closed for unparsed MIPS ABI attributes and macOS saved Hopper databases without a script-capable bridge route.
Reviewed fixes: #2010, #2108, #2513, #2695, #2731, #2771, #2828, #2867, #2868, #2870, #2880, #2919, #2945, #2961, #2981, #2998, #3000, #3002.
Validation: 285 focused source/boundary tests passed. Real Ghidra, IDA, Linux Hopper, and Windows provider runs were not repeated.
BREAKING CHANGE: MIPS targets with unresolved GNU ABI attributes and saved Hopper databases on macOS are refused rather than advertised as supported.
Remove OS-based assumptions about long and non-ASCII string recognition. Require the dedicated fragmented fixture, independently verify full source bytes, and reject gaps, overlaps, incomplete reads and corrupted decoding. Report unrecognized native strings explicitly and retain the actual procedure count.
Verified through 14 verifier regressions and the complete macOS arm64 Hopper 6.6.0-demo CLI/MCP lane with cleanup.
* feat(ghidra): admit 32-bit big-endian PowerPC ELF
Recognize EM_PPC as a provider-neutral `powerpc` family that keeps its ELF
class, byte order, type and flags, and admit ELF32 big-endian ET_EXEC through
a bounded Ghidra lane. ELF64, little-endian, relocatable/shared objects and
EF_PPC_EMB are refused with a specific reason; Hopper and IDA do not gain
PowerPC support, and Windows Ghidra P0 is unchanged.
Ghidra's ELF loader selects PowerPC:BE:32:e500:default for a plain EM_PPC
header, which misreads 750-family code. The lane binds
PowerPC:BE:32:default/default in the profile (language_selection
"admitted-lane-v1"), imports with it and checks the reported language. A
language set through REA_GHIDRA_LANGUAGE_ID still replaces the lane default.
Add a source-owned fixture, fixed-encoding probe, GNU readelf cross-check and
the optional `verify:ghidra:ppc` real-provider lane, plus docs/ghidra-ppc.md.
Refs #1741
* style(ghidra): format shared PowerPC profile integration
* fix(ghidra): preserve caller-selected PowerPC languages
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* docs(i18n): add Greek translations
Give Greek-speaking readers localized onboarding guidance and begin website coverage while keeping incomplete pages on the explicit English fallback.
- Add the reviewed Greek README and place Greek last in every language selector.
- Register Greek website localization and include 206 translated messages.
- Keep incomplete routes nonindexable and replace fixed translation counts with inventory-based coverage.
Refs #2984
* docs(website): improve Greek homepage wording
Move all 18 translated READMEs into docs/readme/ and keep the English entry at the repository root. Update navigation, guide/image paths, website discovery and documentation checks together.
Preserve the contributor's work and current main content. Cover every translated command and HTML destination, and repair the verified Persian heading fragments.
Validation: 203 focused tests on the synchronized head, plus 119 tests and documentation checks on the integration with main fc0dbe7f9; 1,501 local references and 399 authored fragments verified. The complete CI workflow passes. The upstream x64 dylib fixture correction is included; an unchanged-checkout rerun cleared the earlier deadline timing assertion.
Closes#2141
Co-authored-by: Aimal Khan <94559553+aimalexe@users.noreply.github.com>
Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com>
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
* feat(ghidra): admit recordless legacy MIPS32 ELFs with a pinned 32-bit language
Firmware built with pre-ABI-flags GNU toolchains declares EF_MIPS_ARCH_32
with an unspecified ABI field and no PT_MIPS_ABIFLAGS record. The lane
refused them before provider selection, and Ghidra's own header
auto-selection picks a 64-bit R6 language for exactly these files.
Admit ELF32 ET_EXEC headers with EF_MIPS_ARCH_32 or EF_MIPS_ARCH_32R2,
an unspecified ABI field, EF_MIPS_ABI2 clear and no observed ABI flags
record. o32 is recorded as an inferred analysis assumption
(mips_abi_basis), floating-point ABI semantics stay unverified, and the
existing refusals (FP64, NAN2008, ASEs, machine variants, ELF64) are
unchanged. The profile pins -processor MIPS:<LE|BE>:32:default through a
provider-pinned-v1 selection the bridge-reported language_id is checked
against after import; a caller-selected REA_GHIDRA_LANGUAGE_ID still
wins. The lane identity is versioned
elf32-exec-arch32-legacy-pinned-v1.
verify:ghidra:mips gains a recordless legacy round for both byte orders
that strips the ABI record declarations from a real mips32 fixture,
checks REA's interpretation against GNU readelf, asserts the pinned
language after import and the retained assumption limitations through a
snapshot round-trip.
Fixes#1620
* fix(ghidra): require an inspected absence for the recordless MIPS lane
Address review: undefined ABI flags means the record was not inspected,
not that none exists. Only an inspected absence (null) establishes the
recordless legacy admission; a merely unobserved record keeps its
specific refusal. Also document the legacy admission and its pinned
language in the MIPS provider guide.
* fix(ghidra): tighten the recordless MIPS admission boundaries
Address review: a recordless-looking header whose ABI-record inspection
left declaration sources unexamined (carved ELF with a partial section
table and no program-header record) no longer establishes the recordless
admission; it is refused with the inspection limitation retained. An
EF_MIPS_ARCH_32 header that declares an ABI is refused rather than
relabeled through the 32R2 declared-ABI lane.
* fix(ghidra): reject EF_MIPS_ABI_ON32 from the recordless MIPS lane
Address review: bit 0x40 declares the old n32 ABI. A recordless header
carrying it was admitted with the inferred-o32 assumption even though
the header itself declares n32; the recordless shape now requires the
bit clear and refuses it with a specific constraint.
* fix(ghidra): keep recordless relocatable MIPS objects off the legacy lane
The recordless admission now requires ET_EXEC: the lane and its
verification cover executable firmware images, and a recordless ET_REL
object must stay on the declared-ABI path instead of inheriting the
inferred-o32 assumption under an executable lane id. The ABI-record
requirement paragraph is scoped to the declared-ABI lane so it no longer
contradicts the recordless admission documented above it.
* fix(ghidra): name the ET_EXEC scope for recordless objects and probe the legacy lane
A recordless relocatable object is not an incomplete inspection, so its
refusal now names the ET_EXEC scope of the legacy admission instead of
claiming unavailable ABI headers remain unknown. The recordless legacy
verifier round reads the probe pair and applies the full assertMipsProbe
oracle like the standard round, and its report keeps the independent
GNU readelf receipt beside the fixture facts.
* fix(ghidra): retain the concrete ELF32 legacy compiler pin
* fix(ghidra): bind legacy profiles to fully admitted targets
---------
Co-authored-by: BaaRai <baarai@users.noreply.github.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* ci: pin GitHub Actions to full commit SHAs
Replace every third-party `uses: owner/repo@vN` reference in
.github/workflows with the 40-character commit SHA that the tag
currently resolves to, keeping the exact release as a trailing comment.
Each pin was checked with `git ls-remote --tags`: the floating major tag
and the commented release dereference to the same commit. Local
reusable workflows are unchanged.
Re-pin reactivecircus/android-emulator-runner from the SHA of the
annotated `v2` tag object (4c44018e) to the commit it points to
(a421e438, v2.38.0); Actions resolves both, but only the commit SHA is
an immutable reference that tooling can compare with the release.
Update the Windows Ghidra workflow trust-boundary test to require
SHA-pinned setup-node and upload-artifact references instead of the
previous floating tags.
Validation: actionlint v1.7.7 (-shellcheck= -pyflakes=, as in CI), YAML
parse of all workflows, no remaining unpinned third-party `uses:`, and
the workflow boundary/conformance tests.
* style(ci): format action identity assertion
Retain formatting edits produced before the user stopped validation.
Source inspection and git diff --check only. Tests, builds and typechecks skipped by explicit user request. GitHub review writes paused globally.
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(ghidra): admit raw NMOS 6502 analysis
* test(ghidra): require verified POSIX verifier cleanup
* test(ghidra): verify 6502 low-memory raw imports
* docs(skill): document raw NMOS 6502 support
* fix(ghidra): preserve requested Windows raw source paths
* docs(ghidra): include raw 6502 in Windows installation scope
* test(ghidra): verify Windows raw junction rejection through CLI and MCP
* fix(validation): make fixtures portable and preserve explicit Codex home
* fix(ghidra): attest raw mapping state and address space
* fix(ghidra): align diagnostics and regression groups
* fix(ghidra): bind raw attestation to language address space
* chore(ghidra): avoid verifier insertion conflict
* docs(ghidra): align linked raw admission guides
* fix(ghidra): retain 6502 verifier workspace on failure
* test(ghidra): isolate Windows authority doctor fixture
* fix(binary): re-admit Windows raw paths on session reopen
* test(binary): require fresh provider execution after path change
* Constrain 6502 to raw lifecycle and snapshot targets
* Bypass saved reads for unadmitted Windows paths
* Guard direct CLI snapshot replay on Windows paths
* fix(snapshot): gate Windows Ghidra replay on native admission
* fix(snapshot): constrain nested 6502 subjects to raw images
* test: make dyld slice fixtures portable
* test(mcp): expect trace_feature to report references the provider truncated
A provider that reports a capped xrefs read makes the trace truncated with a residual unknown for that target. An explicit complete report and a provider that reports nothing leave the trace as before.
* fix(mcp): report references a provider truncated in trace_feature
Executions gain an optional flag for a capped xrefs read, set by the IDA adapter from the upstream cap. trace_feature marks the trace truncated with a residual unknown for that target only when the provider reports a truncation; a provider that reports nothing keeps its previous result. The trace_feature workflow profile is revised so earlier results are not replayed.
* refactor(ida): report reference truncation whenever the runner observed an xrefs cap
* test(mcp): cover session forwarding, per-target state, native reports and failures of the trace coverage
* test(mcp): format the session forwarding test
* test(ida): expect headless IDA to resolve containing procedures from lookups and segment facts
Cover positive and negative containment, segment boundaries including addresses above 2^53, failures of the lookup and of the segment survey, cancellation, input validation and the admitted program as the only document.
* test(ida): expect feature tracing and address context to work on headless IDA
Run trace_feature, including a capped reference read, and inspect_address_context with and without an explicit document through the IDA session client.
* test(mcp): expect trace_feature availability to follow the search operations it calls
* fix(mcp): derive trace_feature availability from the search operations it calls
* feat(ida): resolve containing procedures and report the admitted program as the document
The modern headless profile resolves an address through the function lookup and, when no function contains it, tells a mapped address from an unmapped one using the database-scoped minimal survey. A survey that is missing, truncated or inconsistent fails the read. list_documents reports the single admitted program. The attached profiles do not declare either operation.
* docs(ida): describe containing procedures, the admitted document and truncated feature traces
* test(ida): verify containing procedures, the admitted document and feature tracing on real IDA
* docs(testing): list the headless IDA workflows the real lane verifies
* test(ida): make each survey failure case isolate a single defect
* test(ida): pin large addresses, segment order, document identity, attached refusals and each trace_feature prerequisite
* test(mcp): expect a repeated reference source to be resolved once in trace_feature
* fix(mcp): resolve the containing procedure of a repeated reference source once
* test(ida): trace the fixture procedure and derive probe addresses from its facts in the real lane
* test(process): format processCaptureLifecycle.test.ts after the merge
* test(ida): declare the alignment each synthetic universal slice offset satisfies
* refactor(ida): move the real lane's feature trace assertions into a helper
* test(ida): expect the real lane's feature trace check to accept references to every matched address
* fix(ida): check each trace reference against its own matched address in the real lane
* test(ida): pin that the new operations reject a document other than the admitted program
* refactor(ida): move the real lane's mapped non-function probe into a helper
* test(ida): expect the real lane's mapped data probe to skip strings inside functions
* fix(ida): try the target's strings in order for the mapped non-function probe
* test(ida): declare the real lane helper types for the verifier tests
* test(ida): expect the real lane to reject a listed string that resolves as unmapped
* fix(ida): fail the real lane when a listed string resolves as unmapped
* feat(ghidra): decode Swift String literals in function dossiers
Swift passes a String as two words. Small literals keep their UTF-8 bytes
inside those words, and immortal large literals keep a count plus a pointer
to their bytes biased by 32. Ghidra creates no data reference for either, so
analyze_function reported no strings for Swift code.
On AArch64 programs that contain Swift metadata, the Ghidra bridge now runs
constant propagation over the function and decodes register pairs where
Swift consumes a String: consecutive argument registers at calls and tail
calls, x0 and x1 at returns, and both registers of one STP. Large literals
must resolve to terminated bytes inside one initialized block that is not
executable. A pair whose registers both come from conditional selects is
skipped, because Ghidra forks each select without correlating the
condition and could otherwise pair words from different cases.
Decoded values are a derivation from the ABI layout rather than observed
references, so they appear in a new optional swift_string_literals field
with their source instruction, registers and raw words, and in a matching
inspect_analysis_view facet. referenced_strings is unchanged.
Refs #2030
* fix(ghidra): bound Swift literal reads and compare decoded strings
Large immortal literals are measured against the remaining heap before
allocation, and a failed fit stays visible as a limitation. The canonical
empty string is kept. Small and large literals now require matching
address nullability, and function comparison treats decoded literals as
string evidence without equating a missing record to an empty one.
* fix(ghidra): budget Swift literal decoding and tighten use sites
Large literals now draw on one heap budget per dossier, measured once
before decoding, instead of a separate check per literal. A literal that
does not fit is listed in a new swift_string_literal_omissions field
with its address, byte length and reason, so consumers read omissions
from structured data rather than from limitation text.
strings_names comparison checks referenced strings and names first and
reports a change there even when Swift literal coverage differs; it
returns unknown only when those facts match and literal coverage is
absent on one side or partial.
A computed jump counts as a tail call only when it has no resolved
target inside the function, so jump tables no longer qualify, and a
return requires a RETURN p-code operation, so traps no longer qualify.
Decoded literals keep control characters, which Swift strings may
contain and JSON escapes. The Swift verifier builds the literal fixture
for arm64 on any Mac host and now also covers the empty string and a
literal containing a tab.
* fix(ghidra): charge escaped JSON and follow copied selects
The Swift literal budget now charges the worst case per UTF-8 byte: UTF-16
text, six escaped JSON characters held in Gson's growing buffer and in the
returned String, and the UTF-8 bytes written to the socket. Literals made
of control characters can no longer exceed the reserved heap.
A jump counts as a tail call only when every resolved target lies outside
the function. The conditional select check follows register copies back
to their definition, so selects copied into argument registers are
recognized like selects written there directly.
strings_names reports a change when both dossiers decoded different
literals at the same use site, even when either side has omissions; it
returns unknown only when no such use site disagrees.
* fix(ghidra): bound literal records and trace selects across branches
Every Swift literal record now pays a fixed cost against the dossier
budget in addition to its text, so many empty or small literals are
bounded as well. The first record that does not fit is kept as an
omission and decoding stops there, which keeps the omission record
itself bounded; omissions now describe small or large literals.
The conditional select check searches every predecessor path for the
reaching definitions, following register copies, so selects made
before a branch into the use block are recognized.
A dossier records Swift literals and their omissions together. Coverage
reads omissions first, and strings_names compares every use site that
neither side omitted: a decoded literal is known where it was recorded,
and an absence is known only in a dossier whose decoding examined every
use site.
* fix(ghidra): stop every flow at the literal budget and route omissions
Ending one flow from the evaluator leaves other queued flows running, so
the budget now records a shared stop that every flow checks before it
examines another instruction. A record that exceeds the budget is routed
by that capacity result rather than by whether text was decoded, so a
small literal over budget becomes an omission instead of a literal.
When coverage is partial, the per site comparison keeps each literal's
representation as well as its value, matching the full strings_names
projection.
* fix(ghidra): keep unresolved Swift tail-call sites unknown
---------
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(server): frame analysis-target content as untrusted data
Strings, decompiled code, page text, scripts, captures and logs returned
by REA originate from the analysed target, which can embed text aimed at
the agent. Nothing marked that content, so a prompt injection inside a
binary or page could steer the agent toward tools that run code.
- Server instructions now state that target-derived tool output is data,
never instructions, and that any request in it to run, install,
download or send something is a prompt-injection attempt to report.
- Tool contracts declare returnsTargetContent. It is on by default; only
seven tools whose results hold REA session state or caller-supplied
values (binary_session, close_binary, export_evidence_bundle,
record_unknown, update_unknown, inspect_adb_client,
inspect_apktool_client) are audited exceptions, so new tools start
marked.
- Successful results of marked tools carry
_meta.provenance = { untrusted_content: true, source: "analysis_target" }.
content and structuredContent are unchanged, and the response budget
accounts for the fixed marker bytes.
Tests cover the encoding and budget accounting, the audited exception
list against the contract catalog, and an in-memory MCP client that sees
the instructions and the marker on a marked tool but not on
binary_session.
* fix(server): mark failed target-content results as untrusted too
The untrusted provenance marker was attached only on the success branch
of ToolResultDelivery, but failures can carry target-derived text as
well: error messages, captured output, and partial observations such as
the terminal frames a failed capture_process_scenario retains in
partial_observation. A host gating later process or device actions on the
marker would have treated an adversarial error payload as trusted.
Every error path in src/server now carries the failing tool's provenance:
- toErrorToolResult and toRecordedPartialErrorToolResult take the tool's
target-content flag explicitly, so each caller is audited by the type
checker; toCallToolResult, toEvidenceToolResult (recording failure), the
response-budget error that replaces an oversized success, and the
Electron partial-evidence path pass contract.returnsTargetContent.
- Admission failures in runAdmittedToolOperation resolve the flag from the
tool name through the new toolReturnsTargetContent, the same predicate
the contract metadata uses; unknown names stay marked.
- The transport's oversized-error replacement keeps the original's marker
and counts it in its budget check, matching the projected error.
- Errors the MCP SDK builds itself (input validation, thrown handlers) are
marked in the transport for target-content tools only.
Error encoding accounts for the marker bytes exactly as success encoding
does. docs/mcp-contracts.md now states that error results carry the
marker. Tests cover marked and unmarked error results, the
capture_process_scenario partial-observation errors in-process and over
the SDK transport, the recorded partial-evidence path, the budget
replacement, the oversized-error replacement, and SDK validation errors
for a marked and a content-free tool.
* fix(contracts): mark session status, close and client probes as target content
The content-free exception set was audited against successful results
only. Now that failures carry the marker too, every exception was
re-audited against its error paths with one question: can the result
include text produced by an external process, provider or device, or
stored content REA did not author?
Removed from TARGET_CONTENT_FREE_TOOLS:
- close_binary: a failed provider cleanup (GhidraClient cleanup error via
createGhidraDiagnostics) projects the launcher's captured stdout and
stderr, which can contain target-derived log text.
- binary_session: capability status reports runtime unavailability as the
provider error message (ProviderAdapterError and similar), and the
provider candidate and Android availability diagnostics include probe
output such as the apktool stderr excerpt.
- inspect_adb_client: the success result carries the version and
installed path parsed from `adb version` stdout, and a failed probe
keeps a stderr_excerpt.
- inspect_apktool_client: the success result carries the first line of
the launcher's `--version` output, and a failed probe keeps a
stderr_excerpt.
- update_unknown: the new revision carries forward the stored question
and domain, which an imported evidence bundle can supply; list_unknowns
and get_evidence_bundle already return that data marked.
The set now holds only export_evidence_bundle (path, byte and record
counts; failures are filesystem, cancellation, or a fixed integrity
message) and record_unknown (the caller's input plus REA-computed ids
and digests; registry errors project only a reason code).
The conformance test pins the reduced set and asserts each removed tool
is marked. Tests that used binary_session as the unmarked example now use
record_unknown. docs/mcp-contracts.md lists the remaining exceptions and
why the removed tools stay marked.
* fix(contracts): include target-content policy in the catalog identity
createToolCatalog() and CATALOG_IDENTITY.tools projected every
client-visible contract field except returnsTargetContent. Because the
untrusted-content framing left schemas, effects, prompts, CLI names and
the package version unchanged, a same-version MCP process started before
the update digested equal to the new server and binary_session reported
it aligned although it never emits _meta.provenance.
Project the flag as returns_target_content in both the digested tool
catalog and CATALOG_IDENTITY.tools, and export createCatalogDigests so a
boundary test proves that toggling the flag on one tool changes the
tools and combined digests.
Add an explicit attached native IDA MCP compatibility profile while preserving the legacy and headless workflows. Keep recorded input identity, complete source observations and the user's open GUI database. Batch unique callee lookups, conservatively prefilter literal strings, preserve Unicode matching and unavailable caller coverage, and require distinct-function switching in native verification.
Verified with Windows IDA Pro 9.3 and the native MCP plugin built from c9d15d98e8a143fc74f6e14037c48f63889b6d6c with a loopback listener. Production CLI/MCP and an unchanged Windows whoami.exe passed. Windows IDA Free was not used. No personal paths or license registration information are included.
Validation: 200 focused tests, static and documentation checks, architecture guards and discovery passed. Historical same-main warm task measurements improved from 675 to 317 ms; response token estimates fell from 16618 to 11509. Fresh-start, overall CPU and memory improvement were not established. Detailed scope is in experiments/analysis-efficiency/ida-native-windows.md.
Co-authored-by: imissapixel <140405963+imissapixel@users.noreply.github.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Co-authored-by: N0zoM1z0 <161784452+N0zoM1z0@users.noreply.github.com>
* feat(ghidra): admit RISC-V ELF analysis through a pinned generic lane
Refs #2063.
Recognize `EM_RISCV` (243) as a provider-neutral family that carries its own ELF
facts, and bound the first Ghidra lane to little-endian `ET_EXEC` images that
declare the double-float ABI, pinned to `RISCV:LE:32:default` or
`RISCV:LE:64:default` with the `gcc` compiler specification.
Ghidra 12.1.4 is not reliable without the pin: its `riscv.opinion` pins
`variant="default"` for the 64-bit constraint and leaves the 32-bit constraint
with the comment that the loader "must choose between 'default' and 'AndeStar'
variants". Measured on the installed release tree, a source-owned generic RV32
ELF with no Andes declaration imports as `RISCV:LE:32:AndeStar_v5:gcc`, while
RV64 resolves to `RISCV:LE:64:default:gcc`. The lane therefore states the
language per width and refuses any caller-selected processor outside it instead
of inheriting the header guess.
Refusals name the declaration rather than falling back to a guessed profile:
big-endian images (Ghidra 12.1.4 ships no big-endian RISC-V language),
non-`ET_EXEC` images, `EF_RISCV_RVE`, `EF_RISCV_TSO`, and every float ABI other
than double, which the pinned generic compiler specifications model. The raw
e_flags bits stay in Evidence as declarations, not as instruction semantics.
Family recognition now has one owner for the ELF-metadata families:
`ELF_METADATA_ARCHITECTURES`, `isElfMetadataArchitecture`, and
`HeaderArchitecture` replace ten `Exclude<BinaryArchitecture, "mips">` sites that
would each have needed a second hand-edited member. `ghidraPinnedLanguage` is
that owner for the import commitment: the session readback previously re-derived
only the MIPS pin, so a profile committing any second pinned language failed with
a language-mismatch error before the bridge started.
Hopper and IDA refuse RISC-V by name, matching the MIPS wording. Caller-supplied
raw-binary entry keeps its existing four-family list; this lane is ELF-header
admission only.
Verification on macOS 15.7.7 arm64, Node 24.18.0, Ghidra 12.1.4 PUBLIC and JDK
21.0.12.1, with the host decompiler built from the release sources for
`os/mac_arm_64`: `npm run check` passes; `npm run test:fast` passes 396 test
files; `npm run verify:test-discovery` discovers all 917 test files exactly once.
Source-owned RV32 and RV64 double-float `ET_EXEC` fixtures complete
`rea analyze --provider ghidra` and `rea decompile rea_riscv_entry`, committing
`RISCV:LE:32:default` and `RISCV:LE:64:default` with `language_selection:
provider-pinned-v1`; the soft-float RV64 shape is refused with the named
float-ABI reason, and the full refusal matrix including the RV32 and
width-mismatch shapes is covered by unit tests. No `verify:ghidra:riscv` lane
script is included in this change; the runs above were performed manually against
the installed engine. Fork pull requests receive no CI check runs, so the
real-provider scope above is the evidence available for review.
* fix(ghidra): bound the RISC-V lane by declared flags and document it
Address review feedback on #2151.
`e_flags` bits outside `EF_RISCV_RVC | EF_RISCV_FLOAT_ABI` are now refused with
the offending value named, so a reserved or newer producer declaration cannot
enter the verified generic lane by falling through the RVE, TSO and float-ABI
checks.
Admission reads `e_type` only, so the refusal text no longer calls `ET_EXEC`
static: a fixed-address declaration does not establish static linkage, and
`PT_INTERP` and dynamic dependencies are not inspected. The session records that
boundary as a limitation rather than inferring a load model.
Add `docs/ghidra-riscv.md` in the shape of the MIPS lane guide (admission
boundary, refusal table, why the language is pinned per class, Evidence fields,
and what the fixtures do not establish), and correct
`docs/firmware-analysis.md`, which still told users that MIPS and RISC-V
decompilation had not been added.
Verification after the change: `npm run check` and `npm run docs:check` pass; the
RISC-V module tests pass (17 lane cases plus 9 parser cases); on Ghidra 12.1.4
with JDK 21.0.12.1 the RV32 and RV64 double-float fixtures still analyze and
decompile with `RISCV:LE:32:default` / `RISCV:LE:64:default` and
`language_selection: provider-pinned-v1`, the soft-float fixture is still refused
by name, and the new linkage limitation appears in the emitted Evidence.
* fix(ghidra): align RISC-V pinning with admission
Reuse the complete lane admission boundary when selecting a language, retain unsigned unsupported flag bits in diagnostics, and derive float ABI metadata from the source declaration.
* test(ghidra): verify RISC-V CLI and MCP workflows
Build source-owned RV32/RV64 fixtures with Clang and LLD, establish independent ELF and symbol facts with LLVM readelf, and verify real provider evidence, instruction bytes, calls, decompilation, schema projections, snapshot replay, refusal diagnostics, source immutability and cleanup.
Verified on macOS ARM64 with Ghidra 12.1.4, JDK 21 and LLVM/LLD 23.1.2. Dynamic linkage, big-endian, Andes and RVE/TSO workflows remain unverified; ISA attributes are not inspected.
---------
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>
* test(ghidra): require pending multi-HI16 coverage in the real MIPS relocatable lane
Post-merge review on #1950: the relocation oracle exposes coverage.multiple_hi16, but the real-Ghidra lane never required it, and Clang's C output never produces a LO16 that resolves several pending HI16s. Ghidra's loader could therefore mishandle that stream while the lane still passed, although most GCC-built kernel modules checked contain it.
The relocatable fixture adds an assembly block with two lui %hi(rea_mips_rel_carry); LLVM's relocation sort then leaves one HI16 pending until a later LO16 that resolves it together with another HI16, including the carry. The lane now requires multiple_hi16: false on the previous fixture, true for all three variants (LE, BE, soft-float module) on real Ghidra 12.1.4.
* test(ghidra): make the MIPS multi-HI16 fixture independent of relocation sorting
Review on #1974: the block used two %lo of rea_mips_rel_carry, a symbol with other %lo uses, so the relocation stream depended on the assembler's HI16/LO16 sort. Clang 21 happened to leave a HI16 pending; Clang 17 (per review) and GNU as 2.45 (checked here: HI@4 LO@8 HI@0 LO@10) pair each HI16 with its own LO16, so multiple_hi16 stays false and the lane fails.
Use a dedicated symbol with two %hi and exactly one %lo. Any assembler that orders each HI16 before a matching LO16 must then emit both HI16s ahead of that single LO16: Clang 21 gives HI@0 HI@4 LO@8, GNU as 2.45 gives HI@4 HI@0 LO@8. The real-Ghidra lane passes for LE, BE and the soft-float module variant with multiple_hi16 true.
* feat(ghidra): admit soft-float MIPS32r2 ELF with integer-register float storage
Soft-float o32 executables (Linux kernel builds, bare-metal firmware) were
refused because their ABI flags record declares no CPR1 and FP ABI 3.
Ghidra's MIPS32 compiler specifications model only hard-float o32, so typed
float/double signatures (DWARF) would be bound to $f12/$f14/$f0 and the
decompiler loses the real dataflow: `return sel ? a : b` decompiled to
`return in_f0;`, and callers passed phantom FPU inputs.
Admit ELF32 ET_EXEC o32 MIPS32r2 targets whose ABI record declares soft-float
with no CPR1, and commit `mips_soft_float_storage: o32-integer-registers-v1`
in their analysis profile only (hard-float profile digests are unchanged).
ReaGhidraMipsSoftFloat.java runs after auto-analysis and, for each function
with a float/double formal type, either keeps the types with the storage
Ghidra's o32 model assigns to a same-size integer (pair alignment, stack
spill), or, for auto-parameters such as hidden struct-return pointers, swaps
float types for same-size integers under dynamic storage. No function is left
on hard-float storage. The report is written last; the provider client
requires it exactly when the profile commits the policy and refuses the
session before any operation otherwise.
Verification: real Ghidra 12.1.4 lane (verify:ghidra:mips) now builds
hard- and soft-float fixtures in both byte orders and checks exact
parameter/return storage through inspect_native_api, caller constants,
snapshot profile and Evidence outcome; test:fast, verify:package and an
installed-tarball soft-float run pass.
* fix(ghidra): model soft-float MIPS as a prototype for every typed callable
Review feedback on #1949: rewriting storage per Function left external
functions and indirect calls through typed function pointers on Ghidra's
hard-float o32 prototype, which still read $f12/$f14/$f0 while the session
reported success. A negative control confirms it: without the change,
`(*rea_mips_soft_callback)(8.5f, 9.5f, sel)` decompiles to
`(*rea_mips_soft_callback)(in_f12, in_f14, sel)`.
ReaGhidraMipsSoftFloat.java now derives a `__rea_soft_float` prototype from
the program's own default o32 model (encoded, float pentries and
float-consuming rules removed, so byte-order-specific aggregate rules and the
hidden struct-return pointer are kept), installs it as a specification
extension, and assigns it with dynamic storage to every non-thunk function,
external function and function-pointer definition whose formal types use
float or double. Thunks follow their targets. Float types are kept
everywhere, so the integer-retype fallback is removed. The script fails
without writing its report if the derived prototype or any reassigned
function still uses an FPU register.
The policy becomes `o32-soft-float-prototype-v1` with a
{functions, definitions} report. The soft-float fixture adds an indirect
call through a typed pointer; the real-Ghidra lane requires its constants
with no FPU inputs and the struct return's `float scale` in $a1.
* test(ghidra): pin ODDSPREG acceptance for soft-float MIPS ABI records
Review on #1949 suggested requiring flags1 == 0 for soft-float, since MIPS_AFL_FLAGS1_ODDSPREG names CPR1 registers. GCC 5.2 kernel modules, LLVM/Rust firmware and Clang -msoft-float objects all record flags1 = 1 with FP ABI soft-float: it is the MIPS32 o32 toolchain default and selects nothing without CPR1. Keep accepting it, say why at the check, and test that soft-float with ODDSPREG is admitted while any other general flag is refused.
* feat(ghidra): admit relocatable MIPS32r2 ELF objects
Linux MIPS kernel modules (.ko) and other ELF32 o32 MIPS32r2 ET_REL objects
were refused by the Ghidra MIPS lane. Ghidra's ELF loader already places each
allocated section and applies REL relocations; this admits ET_REL alongside
ET_EXEC (ET_DYN and PRX stay refused) under its own support lane,
`elf32-rel-o32-arch32r2-standard-v1`, so executable targets keep theirs.
Relocatable sessions carry limitations: addresses are analysis coordinates,
undefined symbols live in Ghidra's EXTERNAL block, and HI16 relocations report
PARTIAL because Ghidra applies them with their paired LO16.
Verification: new `verify:ghidra:mips-relocatable` real-Ghidra lane builds
-c -G0 objects in both byte orders plus a soft-float -mlong-calls module
variant. GNU readelf supplies sections, symbols and REL entries; every
relocation in an allocated section must encode S + A in the loaded word
(S from Ghidra placement, A from original bytes), with HI16/LO16 checked as
the effective address each LO16 forms, and coverage of a shared HI16, a LO16
carry, external code/data and R_MIPS_32 is asserted. The executable lane,
unit suites pass, and on 30 real .ko files 54,972 relocations matched (93 reference empty sections and are not checkable).
* fix(ghidra): verify every MIPS HI16 and archive-typed soft-float externals
Review feedback on #1950:
- The relocation oracle kept one pending HI16, so a second HI16 before the
LO16 that resolves both overwrote the first, and unrelated relocations
cleared it; the earlier site's high half and carry were never compared
with S + A. Every HI16 now waits for the next LO16 of its symbol, which
checks all of them; later LO16s reuse those HI16s, and a HI16 left
unresolved at a section end fails. On 30 real soft-float kernel modules,
24 contain several HI16s per LO16; all 54,972 placed relocations match.
- External callees with float prototypes: the relocatable fixture now calls
powf, whose prototype Ghidra's C library archive applies by name. Every
variant must decompile `powf(1.5,2.5)` without FPU inputs, and the
soft-float module variant must report that __rea_soft_float covered it
(with the external step removed it decompiles as `powf(param_1,param_2)`).
`-fno-builtin` keeps the call from being folded.
Also updates the module variant's expected soft-float policy to
`o32-soft-float-prototype-v1`.
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(ghidra): admit soft-float MIPS32r2 ELF with integer-register float storage
Soft-float o32 executables (Linux kernel builds, bare-metal firmware) were
refused because their ABI flags record declares no CPR1 and FP ABI 3.
Ghidra's MIPS32 compiler specifications model only hard-float o32, so typed
float/double signatures (DWARF) would be bound to $f12/$f14/$f0 and the
decompiler loses the real dataflow: `return sel ? a : b` decompiled to
`return in_f0;`, and callers passed phantom FPU inputs.
Admit ELF32 ET_EXEC o32 MIPS32r2 targets whose ABI record declares soft-float
with no CPR1, and commit `mips_soft_float_storage: o32-integer-registers-v1`
in their analysis profile only (hard-float profile digests are unchanged).
ReaGhidraMipsSoftFloat.java runs after auto-analysis and, for each function
with a float/double formal type, either keeps the types with the storage
Ghidra's o32 model assigns to a same-size integer (pair alignment, stack
spill), or, for auto-parameters such as hidden struct-return pointers, swaps
float types for same-size integers under dynamic storage. No function is left
on hard-float storage. The report is written last; the provider client
requires it exactly when the profile commits the policy and refuses the
session before any operation otherwise.
Verification: real Ghidra 12.1.4 lane (verify:ghidra:mips) now builds
hard- and soft-float fixtures in both byte orders and checks exact
parameter/return storage through inspect_native_api, caller constants,
snapshot profile and Evidence outcome; test:fast, verify:package and an
installed-tarball soft-float run pass.
* fix(ghidra): model soft-float MIPS as a prototype for every typed callable
Review feedback on #1949: rewriting storage per Function left external
functions and indirect calls through typed function pointers on Ghidra's
hard-float o32 prototype, which still read $f12/$f14/$f0 while the session
reported success. A negative control confirms it: without the change,
`(*rea_mips_soft_callback)(8.5f, 9.5f, sel)` decompiles to
`(*rea_mips_soft_callback)(in_f12, in_f14, sel)`.
ReaGhidraMipsSoftFloat.java now derives a `__rea_soft_float` prototype from
the program's own default o32 model (encoded, float pentries and
float-consuming rules removed, so byte-order-specific aggregate rules and the
hidden struct-return pointer are kept), installs it as a specification
extension, and assigns it with dynamic storage to every non-thunk function,
external function and function-pointer definition whose formal types use
float or double. Thunks follow their targets. Float types are kept
everywhere, so the integer-retype fallback is removed. The script fails
without writing its report if the derived prototype or any reassigned
function still uses an FPU register.
The policy becomes `o32-soft-float-prototype-v1` with a
{functions, definitions} report. The soft-float fixture adds an indirect
call through a typed pointer; the real-Ghidra lane requires its constants
with no FPU inputs and the struct return's `float scale` in $a1.
* test(ghidra): pin ODDSPREG acceptance for soft-float MIPS ABI records
Review on #1949 suggested requiring flags1 == 0 for soft-float, since MIPS_AFL_FLAGS1_ODDSPREG names CPR1 registers. GCC 5.2 kernel modules, LLVM/Rust firmware and Clang -msoft-float objects all record flags1 = 1 with FP ABI soft-float: it is the MIPS32 o32 toolchain default and selects nothing without CPR1. Keep accepting it, say why at the check, and test that soft-float with ODDSPREG is admitted while any other general flag is refused.
Move artifact inventory, application projection, runtime reconciliation, and TODO scanner suites into their real boundary lanes. Extract Hopper process/socket cleanup cases while keeping injected session cleanup cases in composition.
Validation: all 56 focused tests passed; discovery finds all 876 test files exactly once. Test assertions and provider verification scope are unchanged.
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>
Return the exact static initializer disassembly before Java projection can
lose field assignments. Retain byte limits and representation markers, and
verify fully folded and mixed initialization through the real CLI/MCP lane.
Closes#1846
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com>
* feat(ghidra): admit bounded PSP ELF through installed Allegrex
Build on merged generic MIPS without weakening its policy. Bind caller-installed processor files into the analysis profile, check them again before launch, force and verify PSP language selection, retain limitations, and extend the public CLI/MCP verifier with source-owned PSP and target-switching checks. Local actual focused Vitest: 102 passed, one existing permission skip. Hosted static and real-provider results are pending.
* fix(ghidra): validate PSPSDK ABI records in PSP admission
* test(ghidra): bind PSP ABI verification to snapshot profiles
* test(ghidra): declare independent PSP ABI reader contract
* test(ghidra): retain the PSP fixture discriminated type
* test(ghidra): contextually type PSP metadata as the MIPS target variant
* test(ghidra): keep carved firmware checks scoped to generic MIPS fixtures
* style(ghidra): format PSP language-override guard
* style(test): format PSP language override regressions
* fix(ghidra): bound PSP extension enumeration before allocation
---------
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Attached IDA mode always uses the legacy 1.4 producer, whose
disassemble_function jumps to the function and whose decompile_function
opens or reuses a pseudocode view whenever an IDA window is active.
Session discovery still advertised may_show_ui: false for every
operation.
Declare the UI effect for the attached operations that reach those
tools: pseudocode, assembly, raw function instructions and the function
dossier. Headless operations and the other attached operations keep
false. A conformance check now compares each declaration with the
upstream tools the operation actually calls.
Fixes#1712
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(build): emit no runtime imports for type-only bindings
Under verbatimModuleSyntax, an import whose specifiers all carry inline
`type` modifiers compiles to `import {} from "..."`, a side-effect import
that still resolves and evaluates the module. 78 compiled modules carried
one, including `import {} from "playwright-core"` in the browser scenario
session, which loaded Playwright's module graph at MCP startup.
Use top-level `import type` for these bindings and enable oxlint's
typescript/no-import-type-side-effects rule so the pattern cannot return.
Refs #1685
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* perf(mcp): load Playwright when a scenario launches, not at startup
The browser scenario and Electron active providers are constructed before
the stdio server serves, and both bound Playwright's launchers at module
load. Resolve `chromium` and `_electron` from playwright-core when a
capture opens instead, as the web module resolver already does.
MCP startup now resolves 1,312 modules (4,902 resolutions) instead of
1,598 (6,076) before reading the first request, and no longer loads
playwright-core. Provider construction, the tool catalog, and capability
inventory are unchanged.
A boundary test spawns the stdio server, initializes, lists tools, and
asserts playwright-core and isomorphic-git were not resolved; the CLI
probe now exercises the real launch path to prove detection.
Closes#1685
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* 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>
Preserve observed terminal-call sites and callee flags across function views, raw bridge responses, Evidence, and snapshot replay. Version analysis profiles so older unqualified results cannot be reused.
Qualify Ghidra function extents without changing external no-return flags or claiming recovery of excluded code.
Validation: 76 focused tests, check:fast, docs:check, formatting, and the full Ghidra 12.1.4 host lane pass. Real CLI and all five stdio MCP function views retain the terminal-call evidence; returning puts calls remain unflagged. The reported stripped Rust ELF and Windows host remain unverified.
Fixes#1658
Check function and label ownership in the existing namespace before renaming. Keep same-address renames idempotent and reject duplicate batch members inside their individual transactions. Extend real-provider namespace checks for single and batch collisions without conflating function and label fixture names.
Validation: real Ghidra 12.1.4 host-native verification passed on macOS ARM64 with JDK 21.0.12.1, including public CLI/stdIO MCP, qualified-name round trips, rollback, source-byte immutability, and complete cleanup. Windows mutation authority remains unchanged.
Fixes#1649
Place the imported control label inside a valid multi-byte NOP instead of suppressing Mach-O function-start metadata. Encode the supported ARM64 and x86-64 instructions explicitly and retain the load dependency so the control remains past the function entry. Assert that Ghidra imports a label and no procedure at the control address before checking name rejection and address containment.
Two real arm64 Ghidra 12.1.4 imports confirmed both fixture strategies produce labels; only default linking retains LC_FUNCTION_STARTS. The full native-value CLI/MCP verifier passed with 349 successful and 52 expected rejected boundary calls, including imported aliases, complete exports, and startup cancellation cleanup. Host ARM64 and Apple x86-64 fixture compilation passed; x86 provider execution remains unverified.
* fix(test): preserve Ghidra interior-label controls on Mach-O
* test: preserve selector semantics and accept signaled fixture exits
* test(javascript): exercise deep returns through the public CLI
* feat(ghidra): admit set_address_name and set_addresses_names
Ghidra sessions rejected the existing naming tools with "operation is not declared by this provider", so agents had to use annotate_native_function, which returns a complete function dossier per rename and cannot name data addresses.
The Java bridge now names a local function entry (namespace-aware, like annotate_native_function) or creates/renames the primary label at any other mapped address, verifying readback. set_addresses_names applies each address in its own transaction and reports per-address booleans. Both are mutating capabilities: excluded from read-only (Windows P0) handshakes and from snapshot replay.
* fix(ghidra): normalize qualified symbol names and reject empty leaves
Share namespace-aware leaf normalization across functions and primary labels. Invalid qualified leaves now return false for naming operations without marking the provider unavailable; function annotations retain typed invalid-input errors.
Extend real CLI/MCP namespace verification with namespaced globals, readback replay, failed-name rollback, mixed batches, and capability checks.
Validation: 75 focused tests and npm run check pass. Ghidra 12.1.4/JDK 21 namespace verification passes on macOS arm64 with unchanged source bytes and complete cleanup. The full host lane passes this workflow before the existing entry-alias fixture fails because the compiler places its interior label at the function entry.
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
* feat(native): admit bounded MIPS ELF analysis through Ghidra
Retain MIPS ELF identity and interpretation through Evidence and snapshots.
Keep provider admission bounded to ELF32 ET_EXEC o32 MIPS32r2, without
implicitly widening Hopper, IDA or Windows P0 coverage. Add source-owned
fixtures and an optional real-provider verifier; real Ghidra verification
remains pending in this contribution environment.
Refs #718
* fix(ghidra): bound MIPS floating-point mode admission
Reject explicit EF_MIPS_FP64 and EF_MIPS_NAN2008 declarations until their
semantics have a real-provider verification lane. Add PSP PRX/Allegrex,
o64/EABI and floating-point-mode refusal regressions. Document that the
header gate is not a full .MIPS.abiflags check and advance the profile lane
identity to invalidate older interpretation bindings.
Validation: 24 standalone checks against the transpiled production profile
module passed; the two new flag cases failed before this change. Clang 17
and LLD compiled the source-owned fixture for both endian targets. The
complete Vitest/typecheck suite and real Ghidra verification remain pending.
Refs morluto/rea#718; PSP support remains follow-up morluto/rea#1330.
* fix(native): validate MIPS ABI records before Ghidra admission
Read bounded ELF32 PT_MIPS_ABIFLAGS and SHT_MIPS_ABIFLAGS records through
the already-open target handle. Preserve declared ABI fields, reject
conflicting/duplicate/truncated records, and keep missing records unknown.
Wire inspection into target resolution before provider selection. Require
compatible declarations for the deliberately narrow Ghidra lane and bind
all ABI fields into profile identity (v3) for snapshot invalidation.
Clang 17 produced eight source-owned fixtures spanning both byte orders
and FPXX/FP32/FP64/soft-float modes with identical ELF header flags. The
new reader and policy distinguish them. 24 focused regression test bodies
passed after TypeScript transpilation using node:test in place of Vitest;
strict isolated TypeScript checking of the reader and target/result types
passed. These are limited checks, not full Vitest/project typechecking or
real-Ghidra verification, which remain pending.
Advances morluto/rea#718. PSP/Allegrex remains morluto/rea#1330 follow-up.
* fix(native): repair MIPS integration boundaries and test resolver behavior
Keep ELF targets out of Hopper Mach-O preparation and scope existing PE fixture types correctly. Move ABI tests into the filesystem boundary lane, use shared cleanup, add resolver regressions, apply the pinned formatter, and register the real MIPS verifier through the existing package-command pattern. Validated with repository-pinned Node/npm, check:ci, actual focused Vitest, discovery and docs:check. Real Ghidra conformance remains pending.
* fix(test): honor caller-selected JDK in real MIPS verification
* test(native): strengthen MIPS Ghidra conformance
Cross-check the source-owned fixture with bounded GNU readelf inspection, compare production ELF/ABI interpretation and independent symbol addresses, and verify effective Ghidra language/compiler identity through existing public operations. Add a source-owned fixed-encoding probe for immediate and conditional branch checks plus global/string byte assertions.
Add 16 assertion/producer-boundary regression cases without widening production admission, changing provider contracts, or introducing proprietary dependencies. Local focused suite: 99 passed, one existing root-permission test skipped. Typecheck, lint/module boundaries, formatting and 700-file discovery pass; full hosted checks and real Ghidra verification remain pending for this checkpoint.
* fix(ghidra): preserve native-tool architecture checks after MIPS integration
The conflict-free upstream merge failed TypeScript compilation because includes on the union of MIPS-only and other architecture arrays requires never. Use equality-based membership without casts or changing accepted native hosts. Extend the existing native-decompiler rejection table with a MIPS executable at a Linux x64 tool path; retain upstream FAT and unknown-header behavior. Hosted verification follows this commit; no success is claimed yet.
---------
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
* feat(mcp): add compact input schema profile for size-capped providers
Moonshot's function-calling API rejects a request while any single tool
parameters schema exceeds its literal serialized size cap, so clients
such as kimi-cli fail every request while the default catalog is
attached (#1484).
Add an opt-in REA_MCP_INPUT_SCHEMA_PROFILE=compact server presentation.
Each advertised input schema is rendered within a 13,312-byte budget,
measured to sit below the observed provider cap: annotation prose
(nested descriptions, examples, defaults, titles) is dropped, repeated
annotation-free subschemas are shared through schema-local $defs/$ref
references the provider validates without expansion, and the three
structurally oversized schemas are advertised in a documented reduced
form. The canonical catalog, tool set, output schemas, and server-side
canonical validation are unchanged; the default full profile is
byte-identical to today's advertisement.
Boundary tests pin every compact schema to the budget, keep them valid
JSON Schema that still accepts every canonical example, hold reduced
presentations at three or fewer, and verify canonical rejection still
guards a permissive advertisement. Verified against the real Moonshot
API through kimi-cli: prompts succeed with the compact catalog attached
where the default catalog fails every request with 400.
* fix(mcp): preserve compact schemas through Kimi conversion
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Keep the Bahasa Indonesia onboarding text from #1508, and fix the formatting
that left it outside the shared README layout: drop the leading space, point
in-page links at the translated headings, align the analysis table, and link
Bahasa Indonesia from every language selector.
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Dev Cee <24010110166@universitasbumigora.ac.id>
The cross-format oracle requires a callee whose recovered control flow
has more than one basic block, and rea_cross_branch provided it through
a symmetric if/else of tail calls. Clang 21 folds that shape into a
conditional negate plus one tail branch, a single basic block, so the
lane rejected a correct Ghidra analysis of the fixture.
Give the branch a loop whose trip count depends on a volatile global:
no compiler can eliminate the backedge, and the callee stays multi-block
at every optimization level on every toolchain.
Share parser modes and collision-free composite keys, enforce architecture guards, and consolidate redundant unit assertions into filesystem, CLI, MCP, and producer-backed workflows. Preserve concurrent upstream boundary fixes and exclude legacy Objective-C class-name bookkeeping labels.
Validation: source/composition suites (3358 passing), focused canonical digest and MCP checks, typecheck/lint, Knip, formatting, architecture guards, and real Foundation archive and Apple legacy simulator CLI/MCP verification. No real Hopper, Ghidra, or IDA verification claim.
BREAKING CHANGE: Objective-C attribute tokens now contain only name and value. Properties report nullable atomicity and readonly facts; class flags, ivar counts, and method requirement flags are nullable when unobserved.
Remove getter, enum, cloned-projection, cache-identity, and obsolete profile assertions along with the shadow JSON Schema engine and dead test-only APIs. Retain producer, filesystem, SDK, lifecycle, integrity, and malformed-format coverage.
Document the priority of real end-to-end workflows, boundary integration, and producer goldens.
Deliver complete Evidence directly through MCP with required nullable analysis profiles and immutable per-server delivery settings. Bind selected environments at construction and require typed cleanup outcomes with retained retry ownership.
Remove test-only routing, duplicate session and snapshot metadata, producer-field repairs, and ambiguous provider arguments. Separate artifact byte identity from occurrence classification and retain typed source-map and partial native facts across CLI, MCP, and verifier consumers.
BREAKING CHANGE: obsolete Evidence envelopes, session aliases, incomplete snapshots, and missing normalized producer facts are no longer accepted.
* fix(ghidra): classify modern Swift manglings in swift type analysis
Categorize modern Swift 4.2+ manglings ($s, _$s) instead of dropping them
in analyze_swift_types. The classifier only accepted the legacy _Tt marker,
so procedure names reported by current providers produced an empty inventory.
Legacy _Tt families and behavior are unchanged. Fixes#1187.
* fix(swift): retain provider aliases and category unknowns
---------
Co-authored-by: Saim Shafique <sx4imm@gmail.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
morluto
Ariel Fikru Avicenna
Pirro Zani
Aimal Khan
LcokHcak
BaaRai
Lluís Ariño
记得吃中饭
Mykola Dementiev
Omisakin Joshua
Ysopwnable
Fernando Pereira
The one -Niu
Aurelio Bernal
Jay-Alexander Elliot
Kaoru0822
Kaoru0822
Syed Rishad
You Wish
sqmagellan
Giridhar
Haythem Sellami
GRAB Racem
Saim Shafique