mirror of
https://github.com/morluto/rea.git
synced 2026-11-10 00:36:35 +00:00
da3bb5aa7b40e56c2fb12fa22dbd9901ea0ea07a
863
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 RRFZ
|
ec7d09176b | fix(mcp): retain partial Playwright trace Evidence on failed imports (#9543) | ||
|
|
22f28644fe |
feat(artifacts): inventory and extract Deno compile executables (#8527)
* feat(artifacts): inventory and extract Deno compile executables Adds Deno standalone artifact support for Mach-O, ELF, and PE carriers, exposing virtual files, entry points, recorded options, and file digests without running the executable. Extraction uses the existing safe output tree, preserves symlink targets as metadata, and assigns indexed paths to unsafe or oversized names. Environment-file values and certificate bytes remain in local Evidence; npm snapshots are reported by length without decoding. The documentation describes supported layouts, parsing limits, and refusal behavior. * fix(artifacts): resolve nested Intel Deno carriers to the enclosing app The backward trailer search stopped at the first confirmed sentinel, so an Intel app storing another Intel Deno executable as a virtual file reported the inner app. Search to the slice start, drop sentinels inside a confirmed payload, and fail closed on side-by-side trailers that differ. Correct the docs on retained environment values and generated regions. * fix(artifacts): keep Deno coverage, remote paths and PE resources truthful Report unknown coverage when the virtual filesystem cannot be listed, keep resolved specifiers when another remote ID is missing, and percent-encode characters in remote paths that a Windows destination would refuse. Reject a directory at the PE resource language level, and correct the guide's description of colliding generated regions. * fix(artifacts): normalize Deno remote escapes so hex case cannot collide Uppercase every percent escape in a generated remote path, so specifiers that differ only in escape case reach the path registry as one name and the second gets an indexed path instead of colliding on a case-insensitive disk. * fix(artifacts): fail closed on truncated enclosing Deno trailers and encode reserved remote path characters The Intel trailer scan rejected an overlong sentinel before reading its marker, so a truncated enclosing app vanished and the nested app was reported. Read the marker first and fail closed when a confirmed overlong trailer encloses another trailer. encodeURIComponent leaves `*` alone and the sanitizer covered only part of the Windows reserved set. remoteDestination now encodes the whole set once as uppercase percent hex; source_path keeps the original. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01KuWHGvzJkrhCqjtEHUwqyH * fix(artifacts): defer truncated Intel Deno trailer failure until enclosing trailers resolve A damaged Intel asset stored before a valid one inside a valid app made the backward trailer search fail before it reached the outer trailer. Hold that failure until the search ends and drop it when a confirmed enclosing trailer holds the sentinel as data. * fix(pe): fail closed when a resource type branch is unreadable beside a found name namedResourceTargets skipped a type branch whose names directory, or a name entry whose string, could not be read. A duplicate RT_RCDATA entry pointing out of bounds therefore vanished once another branch matched, and locatePeResourcePayloads returned only the readable ranges while parsePeResources rejected the same tree. The walk now records unresolved branches. hasPeResourceName still reports false for a tree that never reaches the name, but once locatePeResourcePayloads finds the name it raises a format failure if any branch was left unresolved. * fix(artifacts): skip Intel Deno sentinels that a native segment maps A non-Deno x86_64 Mach-O that stores an Intel Deno executable in __DATA had the asset's trailer promoted to the root application, because the backward sentinel search never checked who owned the bytes it scanned. locateMachOSlice now records the file ranges of every segment other than __SUI and __LINKEDIT (libsui extends __LINKEDIT over the trailer), and the Intel search skips any sentinel inside one of them. A slice whose load commands do not frame stays with native inventory, since ownership cannot be shown there. * fix(artifacts): reject a universal Deno image with an unreadable sibling slice The universal Mach-O locator skipped any slice whose load commands could not be read or framed, and any slice with a FAT range issue. When another slice confirmed Deno, that slice was reported as the whole image, so a sibling declaring a different Deno application went unexamined. locateMachOSlice now reports why part of a slice was not examined (unreadable or unframed load commands, a truncated slice, or the Intel sentinel allowance). locateMachO collects those reasons for every declared and hidden slice and fails closed once any slice confirms a Deno section. With none confirmed, the image stays with native inventory as before. * fix(scripts): supervise the Deno verifier version check Run `deno --version` through the verifier command owner with a deadline instead of raw execFile, and close the owner when the version is refused. * fix(artifacts): keep unconfirmed Deno carriers native and fail truncated Intel trailers An ELF SUI note whose descriptor is too short for the full d3n0l4nd name now stays unconfirmed instead of failing closed; failures after the name matches still fail closed. Every marker-confirmed Intel trailer whose payload runs past the slice is now kept until the backward search ends and fails unless an enclosing payload contains it, so an Intel Deno app missing its last byte reports a Deno failure instead of native inventory. The Deno MCP boundary test now calls open_binary and inspect_artifact through artifactMcpCall, as docs/testing.md requires. * test(deno): drop source-text checks from verifier pin test Keep the wrong-version subprocess check, which covers the pinned release through behavior, and stop asserting on the verifier script's source spelling. * test(deno): close readers in finally in symlink and cancellation tests A failed assertion or read error in either test skipped reader.close(), leaving the file open while temp-directory cleanup ran. Both tests now close the reader in finally, as CONTRIBUTING.md requires. The closed-reader assertion in the symlink test stays after the finally. * docs(specs): add deno-standalone spec, plan and tasks Add docs/specs/artifacts/deno-standalone/ with spec.md, plan.md and tasks.md for the Deno compile executable reader, as AGENTS.md requires for a feature, and list them in the specifications index. * docs: list the Deno compile guide in the documentation sidebar * fix(deno): confirm the d3n0l4nd section before claiming Mach-O damage A __SUI segment alone no longer makes later load-command damage Deno's. A Mach-O whose __SUI segment holds only a differently named section now stays absent and keeps native inventory, while damage seen before or beside a confirmed d3n0l4nd section still fails closed. * fix(deno): keep a slice with an unparsed segment command unread A Mach-O segment command that frames but cannot be parsed may be the __SUI segment, so the slice is no longer treated as fully examined. A universal image whose other slice confirms Deno now fails closed instead of letting that sibling supply the whole inventory, and with no confirmed sibling the image still keeps native inventory. * fix(deno): confirm an ELF SUI note before rejecting its length A SUI note whose descsz runs past its note table was recorded as plain damage before its descriptor name was read, so a malformed Deno carrier fell back to native inventory. The owner and the d3n0l4nd name prefix are now read within the table apart from the claimed extent, and a confirmed note with an oversized extent fails closed. * fix(deno): confirm a truncated __SUI section record by its name A section record cut short after its 16-byte name was skipped, so a damaged Deno carrier fell back to native inventory. The name is now read apart from the rest of the record, and a truncated record that names d3n0l4nd fails closed. The range checks for a confirmed record move into their own helper. * fix(deno): reject nonzero bytes after the last ELF note When fewer than 12 bytes remained in a note table after its last note, the walk stopped without looking at them, so a Deno note was selected with complete coverage beside unaccounted data. The remainder is now read: zero padding is accepted, and nonzero bytes are recorded as damage, which fails closed once a SUI note is confirmed and otherwise leaves the ELF to native inventory, matching the native note parser. * fix(deno): validate FAT slice identity before choosing the search The locator took a universal slice's CPU type from its FAT row, so a row that called an x86_64 slice arm64 skipped that slice's Intel trailer and let the arm64 payload stand for the whole image. Each slice's header is now checked against its row with validateFatSliceIdentity. A mismatch leaves the slice unread, so a universal image whose other slice confirms Deno fails closed. The Intel fixture header now states CPU_SUBTYPE_X86_64_ALL so its FAT rows can match it exactly. * fix(deno): check the Intel trailer marker before judging its length A trailer length word of 2^53 or more was dropped before its d3n0l4nd marker was read, so an enclosing app with that length was skipped and a nested app's range stood for the whole image. The marker is now read first, and a marker-confirmed trailer whose length runs past the slice, safe or not, joins the deferred truncated candidates and fails closed unless an enclosing trailer holds it as data. * fix(deno): count short ELF note tables against the note allowance A note table shorter than one note header made a read of its trailing bytes but counted no note, so many distinct tiny tables could force one sequential read each past the 256-note allowance. Every nonempty table now counts at least one entry, both in the up-front estimate and while scanning, so an image past the allowance records the limit instead of reading every table. The trailing-byte check moves into its own helper. * fix(deno): fail closed on ELF extended section numbering beside a SUI note A zero e_shnum with a nonzero e_shoff moves the section count into section zero, so the section table was skipped and its SHT_NOTE tables were never compared. Record it as damage, as PN_XNUM already is, so a confirmed SUI note fails closed instead of reporting full coverage. * fix(deno): keep undersized Intel trailers until enclosing ownership is known A trailer length word below 8 was dropped before its d3n0l4nd marker was read, so an enclosing app with a length of 0 to 7 was skipped and a nested app's trailer was selected as the root. The marker is now read first, and a marker-confirmed trailer too short to hold it is kept as malformed and fails closed unless an enclosing trailer holds it as data. * fix(pe): treat non-directory RT_RCDATA branches as unresolved A duplicate typeId root entry whose target is a data entry was skipped without marking the walk unresolved, so locatePeResourcePayloads returned the readable named resource as complete. Route non-directory type branches through the same unresolved path as unreadable names directories so the Deno reader fails closed. * fix(deno): fail closed on a zero-fill d3n0l4nd Mach-O section denoSectionRange admitted the declared offset and size of a confirmed __SUI,d3n0l4nd record without checking its section type, so a S_ZEROFILL, S_GB_ZEROFILL or S_THREAD_LOCAL_ZEROFILL section could be read as a file payload while MachoImage reports it as not file backed. Reject zero-fill types with the shared isMachoZeroFillSection predicate before assigning a physical range. * fix(deno): bound the d3n0l4nd Mach-O section by its segment file extent denoSectionRange checked a confirmed __SUI,d3n0l4nd record only against the slice and file bounds, so a section reaching past its segment's fileoff/filesize was still read as a complete payload while MachoImage reports the same section as not file backed. Reject a section outside its enclosing segment's file extent before admitting the physical range. * style(deno): apply oxfmt to the DenoStandaloneLocation embedding union * fix(artifacts): preserve Deno carrier ownership * fix(artifacts): admit Deno parent segments * fix(deno): preserve native segment ownership * test(deno): cover mismatched Mach-O segment widths --------- Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
|
43f06eb1e8 |
feat(native): recover ELF function ranges from unwind tables (#9585)
* feat(native): recover ELF function ranges from unwind tables inspect_elf_unwind_ranges and inspect-elf-unwind-ranges read .eh_frame and .eh_frame_hdr from one local ELF32 or ELF64 image and list the function ranges its FDEs name, with CIE facts, LSDA and personality, the containing section and an exact symbol where one starts at the address. Nothing is executed and no disassembler is needed. Pointer encodings follow the DW_EH_PE rules with raw zero as null, discarded FDEs are counted and skipped, the .eh_frame_hdr search table is compared with the decoded FDEs, and record, row and symbol budgets are bounded. verify:elf:unwind-ranges compares the reported starts with llvm-dwarfdump or readelf on a compiled C++ program. * fix(elf): keep unresolved unwind facts out of reported ranges Skip FDEs whose initial location is indirect and do not locate .eh_frame through an indirect eh_frame_ptr, since pointer slots are not functions. Stop an unbounded header-located scan at a CIE with an unrecognized augmentation, report ignored duplicate PT_GNU_EH_FRAME headers as incomplete coverage, and accept a null personality stored under an indirect encoding. Keep only symbols at decoded FDE starts, cache name observations per string-table offset, and decode names only for the returned page under a fixed 16 MiB text allowance, so repeated or long names no longer grow memory with the symbol count. * fix(scripts): run the ELF unwind verifier under owned process trees Launch probes, compiler drivers, the stripper, the frame dump and the inspection through createVerifierCommands, as verify-binary-hardening does, so a timeout or interrupt stops the whole tree. The temporary directory is removed only after the owned commands have closed. * docs(elf): move ELF unwind ranges guide under docs/guides/native The documentation guide assigns analyst workflows to docs/guides/<domain>/. Move the guide there, place its navigation and directory entries in the Native binaries group, and update links from the docs index, CLI guide, spec and authored skill. The public /elf-unwind-ranges route is unchanged. * fix(elf): give unwind text limits their own recovery and skip indirect tables Throw a dedicated failure when a page's symbol names exceed the 16 MiB text allowance, and map it to a resource error that tells the caller to request a smaller limit rather than raise max_ranges. Read a CIE augmentation with one bounded NUL search and keep it as a byte view, decoded only for returned CIEs, so a long augmentation is never built up character by character. Leave an .eh_frame_hdr search table with an indirect encoding unread: the header is incomplete, the comparison is unavailable, and the table no longer bounds a header-located .eh_frame scan. * fix(elf): resolve header count and table forms and name IFUNC resolvers Leave a non-zero .eh_frame_hdr FDE count under an indirect encoding unresolved, since it is the address of a slot that is not followed, and mark the header incomplete without reading the table. Treat a DW_EH_PE_omit table encoding as an omitted search table before testing the indirect bit, so a header that declares a count without a table stays present with complete coverage, and accept that form in output validation. Match STT_GNU_IFUNC symbols at exact FDE starts after function symbols, with a limitation that the name is the resolver, not the implementation it selects at run time. * fix(elf): require allocated unwind pointer bases * fix(elf): map unwind header relative base * fix(elf): validate unwind load mappings * test(elf): cover mapped nonallocated frame agreement --------- Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
|
f21fda0fc0 |
feat(elf): report a Linux kernel image's version banner and embedded build config (#9573)
* feat(elf): report a Linux kernel image's version banner and embedded build config inspect_kernel_build_config and inspect-kernel-build-config read one local vmlinux, raw arm64 Image, gzip bzImage or zImage, or configs.ko and report its Linux version banner and the IKCONFIG .config embedded between IKCFG_ST and IKCFG_ED, without executing it. A config is present only when a gzip member ends exactly at IKCFG_ED, passes its CRC-32 and ISIZE checks and parses as .config text; framing that fails is reported malformed, never absent. bzImage gzip payloads are inflated within a budget and other codecs are named as unsupported, using codec detection now shared with the kernel module parser. The config search, inflate sizes and option counts are bounded, scans honor cancellation, and target-derived text is marked untrusted. * fix(elf): reject a gzip member whose header checksum fails When FHCRC is set, compare the low 16 bits of the header CRC-32 before inflating, so a damaged header no longer yields a present config or an inflated bzImage payload. * fix(elf): detect the bzImage codec inside the declared payload Codec detection now reads the same payload_offset/payload_length slice that is inflated, so bytes past a short payload cannot name its codec. * fix(elf): report a framed config that does not decode as malformed A damaged or cut-off deflate stream between IKCFG_ST and IKCFG_ED was skipped and could end as an absent config with complete coverage. It is now malformed with config_gzip_invalid, and a later valid config still wins. * fix(elf): scan config lines in place and accept long string values Config text is read one line at a time instead of split into an array, and a quoted value is validated before it is decoded once. The 4096-character string cap is removed: Kconfig has no such limit and max_config_bytes already bounds every line, so a long CONFIG_EXTRA_FIRMWARE list is no longer reported as malformed. * fix(elf): let a live cancellation reach the kernel config parser The parser ran synchronously, so an abort raised during parsing could not change the signal before it returned. Inflation now runs on the zlib thread pool and each scan yields to the event loop once per MiB, so the existing cancellation checks observe a real abort. * refactor(elf): carry the refused kernel config budget as a typed fact Budget failures are now KernelBuildConfigBudgetFailure with a knob field, and the provider builds its recovery advice from that field instead of searching the diagnostic message for an option name. * docs(elf): move the kernel build config guide under docs/guides/native The documentation guide assigns analyst workflows to docs/guides/<domain>/. Move the guide there, place its navigation and directory entries in the Native binaries group, and update links from the docs index, CLI guide, spec and authored skill. The public /kernel-build-config route is unchanged. * fix(elf): yield while checking one long config line A config made of one long quoted value never reached the per-line yield, so the control-byte check, value validation and SHA-256 ran in one synchronous block. Each now works one MiB window per event-loop turn, and the line patterns match only a bounded name prefix instead of the whole value. * fix(elf): return paged config entries that do not pin the decoded text Entry names and values were V8 slices of the whole decoded config, so a one-entry page kept the full text alive. Entries in the returned page are now standalone copies, so retained memory follows the page size. * fix(elf): yield while kernel marker searches find nothing IKCFG_ST, IKCFG_ED and banner searches used one Buffer.indexOf over the rest of the image, so a large image with no marker never yielded. They now search overlapping one MiB windows and yield after each window without a match. * fix(elf): preserve long kernel config names * fix(elf): frame kernel configs by gzip member boundaries --------- Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
morlutoandhhang
|
24053f7ee2 |
fix(desktop): preserve escaped xprop text (#9652)
* fix(desktop): preserve xprop escaped window text * test(mcp): cover source fitting and omitted delivery budgets * test(javascript): distinguish package scope in route fixtures * test(apple): separate identity and refusal fixture groups --------- Co-authored-by: hhang <169027046+xu91102@users.noreply.github.com> |
||
|
morluto
|
9e0bc71e78 |
fix(comparison): retain integrity conflicts for duplicate paths (#9647)
fix: retain integrity conflicts for duplicate artifact paths |
||
|
|
3230f50b16 |
feat(sqlite)!: page selected rows with a stated order and row offset (#9331)
* feat(sqlite): page selected rows with a stated order and row offset readRows read `NOT INDEXED LIMIT ?` with no ORDER BY and reported order: "unspecified", so an agent could not reach rows past row_limit deterministically: raising row_limit hits the fixed 16 MiB reply budget, and the sampled order depends on whichever index SQLite happened to scan. Read the window in a unique, stated order and expose the position: - rowid tables order by rowid unless all of rowid, _rowid_ and oid are declared columns, in which case the order stays unspecified. - WITHOUT ROWID tables order by the primary-key columns, reusing each key's DESC flag and declared collation from pragma_index_xinfo so the plan keeps index order instead of materialising a sorter. - rows.order, rows.row_offset and rows.next_row_offset report the window; next_row_offset is null once the window reaches the end. - row_offset (CLI --row-offset) requires table and defaults to 0. A non-zero offset is refused when the table declares no usable order, because the skipped prefix would be arbitrary. The schema now checks next_row_offset against truncated and the offset so a worker reply cannot claim a page it did not read. One attribution fix came with it: adding ORDER BY text can push an over-wide projection past SQLite's SQL-text limit, which fails as SQLITE_TOOBIG before the result-column limit is reported. Attribute that case to column capacity too, so a 250k-column request still reads as capacity rather than a byte budget. Verified: npm run check:fast, npm run format:check on the touched files, npm run verify:module-boundaries, npm run docs:check, and 196 tests across the sqlite boundary, application and CLI/MCP suites. A 1002-row inspection and in-memory EXPLAIN QUERY PLAN checks confirmed the WITHOUT ROWID path adds no sorter. Full check:pr and the complete test suite were not run. * fix(sqlite): resolve the rowid alias and keep paging claims reachable Three defects in the row paging added alongside row_offset: - readRowOrder emitted ORDER BY rowid for every rowid table, but a table that declares a rowid column shadows that spelling, so the sort used the stored column and could repeat or drop rows across pages. Order now names the first alias the table leaves unbound, read from raw name bytes so an unrelated malformed identifier is not decoded just to choose an order. - The inventory projection could hide the primary key a WITHOUT ROWID table orders by, silently degrading a paginated read to unspecified order. The order key is now read from the table itself rather than from the columns the caller asked to see. - A truncated table with no stated order reported a next_row_offset that the row_offset rule then refused. It is null there, and the reply limitation no longer claims an unspecified order for windows that have one. rowEvaluationFailure also probed the first rows of the table instead of the requested window, so a failure confined to a later page was attributed to columns that evaluate fine there. * test(sqlite): assert the concatenated pages The paging regression checked only counts, offsets and flags, so a second page that repeated the first page's rows would still pass. Assert the values read across all three pages instead. * fix(sqlite): decode row keys in the database encoding readRowKeyFacts decoded primary-key names as UTF-8 and probed a fixed hidden set, so two supported schemas could not be paged at all: - A UTF-16 database stores key names as UTF-16 bytes, so "k" decoded to NUL-padded text and ORDER BY named an identifier that does not exist. Pass the detected encoding, as readColumns and the index-metadata path already do. - A virtual generated column named rowid (hidden=2) was absent from the shadow check, so ORDER BY rowid evaluated the generated expression instead of the rowid and could sort non-uniquely or fail outright. Include generated columns in the alias probe. Also stop decoding primary-key names for rowid tables: their ORDER BY names an alias, never a declared key, so an unselected malformed key name could abort an otherwise readable selection. The page-concatenation assertion now narrows SqliteValue before reading .value, which the null/blob arms do not carry; this was failing the Static checks typecheck job. * fix(sqlite): order WITHOUT ROWID rows by the index's own key sequence `readRowOrder` filed `pragma_index_xinfo` key terms into a map keyed by column name, so a valid composite key that repeats a column — `PRIMARY KEY(a COLLATE BINARY, a COLLATE NOCASE DESC)` — collapsed to one term. The generated clause then overstated uniqueness: it omitted the BINARY term, so `A`/`a` ties survived, and the `row_offset` paging advertised a `primary_key` order that the actual key does not establish. Build the terms from the index rows in `seqno` order instead, keeping repetitions, and bail to `unspecified` if any declared key column never appears in the index. The declared collation stays explicit so a caller column's own collation cannot re-sort the window. Verified against a real SQLite database that the repeated key is accepted and that the old clause orders A, a, B while the index orders A, B, a. * fix(sqlite): accept a null next offset on unordered truncated rows readRows already withholds next_row_offset when a table shadows rowid, _rowid_ and oid and declares no primary key, because every non-zero offset is refused there. The reply schema still required the offset whenever truncated was set, so the provider rejected its own worker result and the first sample never reached CLI or MCP. Require the offset only when the order is stated. * fix(sqlite): probe row keys only for ordinary tables and name row_offset readRowKeyFacts ran for a selected virtual shadow table before readRows could reject it, so an unavailable module surfaced as SQLite's "no such module" instead of the selection refusal. Probe the key only when the selected kind is an ordinary table and leave the order unspecified otherwise, since no rows are read for that kind. The unordered-offset refusal also reported the generic selection reason, which callers see against table. Give it its own reason so the failure is attributed to row_offset. * feat(sqlite)!: page selected rows with a stated order and row offset Selected-row inspection currently reports only rows.order: "unspecified" and offers no way to reach rows past the first sample deterministically. Main also refuses non-zero row_offset today, so nothing relies on that omission; this change makes the stated order and the offset the contract, reports the next offset the caller feeds back, and keeps the existing byte, heap, statement and deadline limits. BREAKING CHANGE: rows.order becomes rowid | primary_key | unspecified, and rows gains required row_offset and next_row_offset members. A client that validates selected-table output against the previous schema (where order was the only permitted value) rejects a now-valid reply. Migration: accept the two new members and treat order: "rowid" | "primary_key" as the stated order the offset applies to; a null next_row_offset means the window reached the end. * docs(sqlite): say a null next offset also means no stated order The selected-rows limitation described next_row_offset as null only when the window reached the end. An unordered table whose row_limit truncates the sample also publishes null, because an offset there would name a request the row-offset rule refuses, so the Evidence read as if the sample were complete. State the second cause and pin it in the existing regression. * test(sqlite): drop the limitation wording assertion and state the unordered exception The boundary test pinned the sentence a caller reads in `limitations`, so a harmless rewrite of that text failed a paging test whose real subject is `order`, `truncated` and `next_row_offset`. docs/development/testing.md asks for facts over frozen explanatory prose, and the three structured expectations already fail if the paging fact changes. The tool contract promised that retained rows always have a unique order and can be paged. A rowid table that shadows rowid, _rowid_ and oid and declares no primary key has no such order: it returns order=unspecified with a null next offset and refuses a non-zero row_offset. Say so in the description so callers do not plan a continuation the tool always refuses. * docs(sqlite): say unordered tables cannot be paged The description told callers that a truncated table without a stated order needs "the table read without offsets", but omitting row_offset starts at zero again: with an unchanged row_limit the next call repeats the first sample instead of continuing past it. Say that these tables cannot be paged and point at row_limit for a larger first sample. * style(cli): wrap the native data type path description The `inspect-native-data-type` argument description line is 82 columns, over the repository's 80-column `.oxfmtrc.json` limit, so `npm run format:check` fails on main and on every branch merged from it. It came in with #9465 and the following commits to this file did not touch it. Reformat the one call the formatter objects to; no behaviour change. * fix(sqlite): bound rowid alias comparison by name byte length readRowKeyFacts decoded every column name before testing it against three short aliases, so a table declaring a multi-megabyte name could exhaust the worker heap even when the caller projected the column away. Reject names that cannot be an alias by their byte length, accounting for the database encoding, before building any string. --------- Co-authored-by: 2SinceJuly <202121119+2SinceJuly@users.noreply.github.com> Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
|
97b5503f39 |
feat(apple): join IPS images to admitted Mach-O slices (#9570)
* feat(apple): join Apple IPS frames to a selected Mach-O or dSYM slice inspect_apple_ips reported every image UUID as unknown and stated that joined frames were unobserved. Accept an optional join_path, read its LC_UUID and __TEXT vmaddr with the existing bounded stable-artifact reader, and match by canonical UUID plus architecture. A frame joins only when exactly one slice carries the image UUID and an agreeing architecture, so arm64 and arm64e never cross, all-zero UUIDs never join, and a conflicted or unreadable artifact stays typed instead of guessed. Joined frames report the slice __TEXT vmaddr plus the recorded imageOffset as a static address, which is the value atos resolves for the same report. Symbolication is still not performed. * refactor(apple): remove superseded join integer type import * fix(apple): share FAT admission and preserve completed join facts * fix(apple): bind standalone joins to admitted source identity --------- Co-authored-by: 2SinceJuly <202121119+2SinceJuly@users.noreply.github.com> Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
Matt Van HornandMatt Van Horn
|
645735fc0f |
feat(windows): inspect Shell Link (.lnk) shortcut files (#7859)
* feat(windows): inspect Shell Link (.lnk) shortcut files Add inspect_windows_shortcut / inspect-windows-shortcut to read one local Shell Link without resolving or executing the target. The report keeps header flags and FILETIMEs, IDList items, LinkInfo volume and network names, StringData, ExtraData including Tracker droid facts labelled derived, and derived observations such as argument length and a minimized ShowCommand. * fix(windows): tighten Shell Link string reads and coverage - Decode counted StringData over its full CountCharacters payload. - Locate LinkInfo string terminators before charging max_string_bytes and before decoding; an unterminated string is a coverage issue. - Mark an IDList without a TerminalID and an out-of-range LinkInfoHeaderSize as partial coverage. - Report arguments_present and argument_length as null when HasArguments is set but the field could not be read. - Derive target_extension from the joined base path and suffix. - Keep ISO dates for FILETIME and UUIDv1 timestamps before 1970. - Retain the completed file snapshot on parse failures. - Reject an empty CLI path before normalization. - Select the shortcut contract from its family in focused tests. - Fixture StringData no longer writes a NUL into the counted payload. * style: format docs index and Shell Link boundary type Apply oxfmt to the two files format:check reports so Static checks passes. * fix(windows): reject network share paths for Shell Link reads inspect_windows_shortcut is advertised as offline, but the Windows path schema admits UNC paths, and lstat on one can contact an SMB server and offer the caller's credentials. Reject UNC, extended UNC, and other host-prefixed device paths in the provider before any filesystem call. Extended drive paths such as \\?\C:\x stay allowed. A mapped network drive cannot be told from a local drive by its path, and the docs say so. * fix(windows): decode Unicode Shell Link file-entry names File-entry items with bit 0x04 of the type byte, such as 0x36, store a UTF-16LE name, but the parser always read it as Latin-1. The read stopped at the first zero byte, so report.exe became r and an IDList-only shortcut reported the wrong target_extension. Read those names as UTF-16LE. * fix(windows): read volume labels by exact Shell Link discriminator VolumeLabelOffset 0x14 selects VolumeLabelOffsetUnicode and is not itself a label offset, and that Unicode field exists for no other value. The parser treated any offset of 0x14 or more as having both, so a Unicode-only label such as "OS" also produced a false ANSI label "O", and an ANSI label stored past 0x14 had its own bytes read as a Unicode offset. Branch on the exact value in parseVolume, the one place that interprets it, and report a VolumeID too short to hold the selected field. The fixture wrote both encodings for a Unicode volume, which hid the false label, so it now writes one. * fix(windows): limit network share rejection to Windows hosts POSIX hosts read repeated leading slashes as one local root, so an accepted selection such as ///tmp/shortcut.lnk is a local file there. Apply the host-prefixed path check only on Windows, and take the platform through the provider constructor so the Windows path forms stay covered on every test host. * fix(windows): compare shortcut icons only against full paths A stored IDList filename or relative path cannot show that an icon at a full path is elsewhere, so the comparison reported a false difference. Report null unless the icon and a recorded target are both full paths, and assert the missing TerminalID issue by code and offset instead of its message. * fix(windows): keep unread ExtraData blocks unknown A walk that stopped before or inside ExtraData still reported the blocks that HasExpString and HasExpIcon expect as absent, though they were never checked for. Report a block as absent only after the TerminalBlock shows the section was read in full. Without one the block stays unknown and coverage stays partial. * fix(native): ignore LNK network provider type unless ValidNetType is set CommonNetworkRelativeLink NetworkProviderType is only meaningful when the ValidNetType flag (0x2) is set. Return null otherwise. * fix(native): accept any ExtraData TerminalSize below 4 as the terminal MS-SHLLINK requires TerminalSize to be less than 4, not exactly 0. Treat 1, 2 and 3 as the terminal block instead of a malformed block. * fix(native): derive shortcut target facts from the stored environment target Include the EnvironmentVariableDataBlock literal target among the target candidates for target_extension and the icon comparison. Variables are not expanded and the target is not opened. * fix(windows-shortcut): keep unexpanded variable paths unresolved in icon comparison A stored environment target such as C:\Users\%USERNAME%\app.exe could be compared as different from C:\Users\Alice\app.exe. Variable-bearing paths are never expanded, so the icon comparison now returns null unless a fixed target matches the icon exactly. * test: drop redundant ok assertion before typed error guard The testing guide asks for the typed error to be thrown before asserting a successful Result value, and says a preceding .ok assertion adds no coverage. * fix(windows): point out-of-range LinkInfo offsets at their field A LinkInfo, VolumeID or CommonNetworkRelativeLink offset that points past its structure was reported at base + offset. With an offset such as 0xffffffff that position exceeds uint32, so the report failed its schema and the service discarded the whole partial result. Report such issues at the offending offset field instead, which keeps every issue offset and coverage.failing_offset inside the file. Cover each LinkInfo offset field in the parser tests and the LocalBasePathOffset case through the stable filesystem service. * fix(windows): require RFC 4122 variant before deriving Tracker UUID facts A version nibble of 1 alone does not identify a version-1 UUID. Also require the clock_seq_hi_and_reserved top bits to be 10, so other GUID layouts keep their raw value and derive no MAC or timestamp. * fix(windows): ignore DeviceNameOffset fields unless ValidDevice is set MS-SHLLINK 2.3.2 says DeviceNameOffset and DeviceNameOffsetUnicode are ignored when the ValidDevice flag is clear. parseNetwork read both regardless, so garbage in those fields became device-name facts or malformed_structure issues. Read the flags once and only read the device name fields when ValidDevice is set. * fix(windows): keep unsupported Tracker headers as raw data Check TrackerDataBlock Length (0x58) and Version (0) before decoding. A block with any other header is retained as raw bytes with a malformed_structure issue, so it no longer yields MAC and timestamp facts under complete coverage. * fix(windows): compare shortcut icon paths by normalized location Collapse "." and ".." segments and repeated separators in the icon comparison as text only, never opening a path. ".." stops at the drive or UNC share root, and \\?\ and \\.\ paths stay verbatim. Join LocalBasePath with CommonPathSuffix as the target and use the bare base only when the suffix is empty, as the network share branch does. Read the selected shortcut with lstat and open only, so a rejected symlink's destination is never resolved. * fix(windows): keep the UNC share in the root after repeated separators normalizePath took the host and the next split part as the UNC root, so \\server\\share\..\app.exe kept an empty share and let .. remove the real share. Drop empty parts before picking the root so paths on different shares stay different. * fix(cli): parse windows shortcut limits as decimals The four inspect-windows-shortcut limit options now go through cliDecimalField like the other inspection commands, so values such as 0x10 are rejected and cliDecimalArguments.test.ts passes. * fix(native): reject linked parents of windows shortcuts A parent directory symlink or junction pointing at a UNC share was followed by the first lstat, since the network check reads only the path text and the symlink rejection covers only the final component. On Windows, lstat each ancestor from the drive root down and reject the path at the first link, before anything follows it. * docs(native): move Windows shortcut guide and add its feature spec Move the Shell Link guide to docs/guides/native/windows-shortcut.md and list it under Native binaries in the site navigation, keeping the /windows-shortcut public route. Update README, translated README, CLI page and authored skill links to the new physical path. Add docs/specs/native/windows-shortcut/ with spec.md, plan.md and tasks.md, as required by docs/development/documentation.md. * fix(native): admit the windows shortcut parent walk On Windows the ancestor link walk ran before the provider's resource admission, so close() could finish while the walk kept calling lstat. Run the walk inside the same admission as the read, and check the caller's signal before each ancestor lstat. * fix(native): walk the effective parents of windows shortcuts The ancestor link walk checked every prefix of the path as written, so C:\alias\..\real\a.lnk was refused when C:\alias is a directory link, although Windows removes alias\.. before opening the file. Walk the ancestors left after removing . and .. components, keeping the caller's spelling, never climbing above an absolute root, and leaving \\?\ paths as written since Windows does not normalize them. * test(native): assert the linked parent as a structured fact The parent link regression matched the phrase "junction, which" in the refusal message, so a wording edit would fail it. Report the linked ancestor as observed.linked_ancestor on the input issue and assert that fact instead of the explanatory text. * test(native): bound the shortcut close gate and release it in finally Race the parent walk lstat acknowledgement against inspection settling with awaitAcknowledgement, then release the gate and join the inspect and close work in finally so a failure reports its cause instead of waiting for the Vitest timeout. * fix(native): walk Windows shortcut parents lazily windowsAncestors built and kept every parent prefix before the first lstat, so a path with thousands of components allocated memory quadratic in its length. It now keeps only the normalized components and yields each parent from a running prefix as the walk reaches it. --------- Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com> |
||
|
吴佳浩Albenandmorluto
|
25885b05c0 |
feat(v8): inspect V8 code cache containers and retained byte patterns (#8993)
* feat(v8): inspect V8 code cache containers and their retained byte patterns Bytenode-compiled `.jsc` files — the V8 code cache container that `vm.Script`, `vm.compileFunction` and bytenode 2.x write, and that Node single-executable applications and Electron apps ship — had no format-native reader. `inspect_binary_layout` shows the surrounding ELF/PE and `inspect_artifact` inventories the file, but neither decodes the V8 structures, so an agent could not tell which V8 line produced the file, could not learn the container is metadata rather than source, and had nothing to point at a version-pinned decompiler. Issue #8543. `inspect_v8_code_cache` joins the format-native family with `inspect_python_pyc`, `inspect_erlang_beam` and `inspect_kernel_module`: one explicit local file, bounded parsing, no execution, no installation, no child process. **The container is version-dependent, so the reader reports roles with confidence instead of assuming an ABI.** All eight header words are reported with byte offset, observed value, the role this reader believes the word has, and whether that role is `confirmed`, `inferred`, or `unconfirmed`. Only three facts gate admission: the header size this reader pins, the V8 code cache magic high half `0xc0de` (the low half is `ExternalReferenceTable::kSize` and varies per build, so it is a build fingerprint rather than a constant), and a declared payload length that reaches exactly the end of the file. I derived the field semantics from the producer rather than assuming them, and two findings changed the design: - **The word V8 calls the source hash holds the source length on the measured producers.** A 54-code-unit source stored `54`, and sources with the same code-unit length but different content stored the same value, including a source whose UTF-8 length was 154. The reader reports it as `source_length` and the docs say that this name is what the measurement supports, not a claim about what another build stores there. - **`vm.compileFunction` sets a high bit in that word.** The reader masks it out and reports the producer family as `script` or `compile_function`, so the source length stays correct for both producers. Embedded strings are reported as **byte-pattern observations with absolute byte ranges**, labelled as such in the field name and in the limitations. A reported sequence is not proof V8 retained it as a string, and a missing literal is not proof it is absent. **Fail-closed boundaries, all covered by tests:** an unknown magic is a typed `unsupported_target` failure because the file is not a V8 code cache; a short header or a payload-length disagreement is a typed `invalid_request` failure. Neither yields a partial parse and neither records Evidence. Non-goals are explicit in the tool description and the docs: the payload is serialized V8 heap data, not source; recovering logic needs a version-pinned View8-style decompiler that this tool neither provides nor depends on. Fixtures are generated from the running V8 instead of committing a version-pinned binary, so the tests exercise the real producer and stay valid across Node upgrades. - Semantic owner and earliest changed stage: format-native V8 code cache container decoding. - CLI and MCP/tool-catalog contract: one new `artifact-provider` tool and one new `inspect-v8-code-cache` CLI command, registered in `TOOL_CONTRACTS`, the `artifact` language family, `TOOL_EFFECTS`, the capability declarations and services, and the capability inventory. `docs:check` passes, so the generated catalog agrees. - Provider, bridge, target-format, or platform compatibility: no provider runs and no external tool is required, so the tool is available on every supported Node host. - Evidence, artifact, provenance, or reconstruction contract: complete Evidence with the payload byte range as its location; no artifact format, reader, or reconstruction change. - Generated metadata, package, or installation impact: catalog and document facts regenerated by the existing checks; no new dependency. Expected effects: - [x] Target content is returned as bounded structured facts. - `src/v8/V8CodeCacheParser.test.ts` (13 cases): real-producer header facts, per-field role and confidence, source length for both producers, an observation's absolute offset verified against the file bytes, the limitation text, foreign magic, payload-length mismatch, short header, missing payload bytes, a different build constant still admitted, and the observation budget. - `tests/boundary/mcp/v8CodeCacheMcp.test.ts` (4 cases): advertisement and annotations, header facts over MCP, relative-path rejection, unsupported magic, malformed container. - `tests/acceptance/v8CodeCacheCli.test.ts` (4 cases): the public CLI, relative-path resolution, byte-range observations, and the typed failure. - `npx vitest run src/contracts src/v8 src/application/v8CodeCache src/domain/v8` — 242 passed - `npx vitest run tests/boundary/mcp/v8CodeCacheMcp.test.ts tests/acceptance/v8CodeCacheCli.test.ts tests/boundary/filesystem/productCatalog.test.ts tests/acceptance/tools/localToolContracts.test.ts` — 20 passed, 4 skipped - `npx tsc --noEmit -p .` — exit 0 - `npm run build` — exit 0 - `npm run lint` — 0 errors - `npm run knip` — exit 0 - `npm run docs:check` — exit 0 - `npx oxfmt --check src tests docs` — exit 0 Gap: the header table is measured on node v24.22.1 / V8 13.6.233.17-node.58 only. No other V8 build is available here, so a different build's field order is untested; that is exactly why only three facts gate admission and the remaining roles carry a confidence level rather than being asserted. Additive. No existing tool, schema, or behaviour changes. The reader never executes the selected file, never launches a process, and never accesses the network; its audited effects are `mutatesSession: true` only, matching the other static format readers. Human owner: @XiaoBinGan - [x] I reviewed the final diff and relevant existing code, understand the change and its edge cases, and take responsibility for the complete PR, including AI-assisted work. - [x] I checked the evidence above, confirmed it supports the current change, and recorded meaningful validation gaps. - [x] I declare this PR ready for initial maintainer review. - [x] This PR addresses a concrete problem or an agreed enhancement. - [x] Material scope and authority questions are resolved, and any required maintainer scope decision covers the final diff. - [x] The PR has one focused outcome and the title follows `type(scope): outcome`. - [x] Related issue: #8543. - [x] Tests cover changed observable behavior and meaningful failure paths. - [x] Owning docs, contracts, and generated metadata are updated where needed. - [x] I checked the final diff for secrets, unrelated cleanup, and unsupported claims. * chore: retrigger title and readiness checks after the title was corrected * fix(v8): bound the payload scan and use the repository path conventions Review feedback from @greptile-apps on four points. **Bound the synchronous payload scan.** The scan is one synchronous pass over the payload, so the 512 MiB ceiling allowed a single request to hold the process for several seconds. The ceiling is now 64 MiB, matching the pinned CPython pyc reader, and the default is 32 MiB. The scan stays synchronous, like the BEAM and ELF note readers in this family, but the budget that bounds it is no longer large enough to matter. The new `max_file_bytes` ceiling and default are documented. **Use the absolute-path schema.** `path` now uses `absoluteLocalPathStringSchema` instead of `localPathStringSchema`, matching `inspect_erlang_beam`. A relative name previously passed validation and was resolved against the server's working directory, so `module.jsc` could select a different file than the caller meant. The rejection is now at the schema boundary. **State the extraction step.** The handoff in `docs/v8-code-cache.md` now says a cache embedded in a single-executable application or an Electron app must be extracted first, because this reader requires the magic at byte zero and returns `unsupported_target` for the surrounding ELF/PE executable. **Load only the owning contract group in tests.** Both tests now import `V8_CODE_CACHE_TOOL_CONTRACTS` directly instead of resolving the tool through the full catalog, so a focused test no longer constructs every other tool's schema. 21 cases pass across the parser, MCP boundary and CLI acceptance files; `tsc` 0; `build` 0; `knip` 0; `docs:check` 0; `oxfmt --check` 0. * fix(v8): select the tool contract by name in the focused tests Reverts the catalog-position selection I added for a review suggestion. This repository has a lint rule (`rea(no-positional-tool-contract)`) that forbids selecting a tool contract by catalog order, and `toolContract("inspect_v8_code_cache")` is how every neighbouring focused test selects one. The positional form I tried introduced two lint errors, so the named form stays. The review concern was that resolving through the full catalog constructs every other tool's schema. That cost is real but it is the repository's chosen trade: catalog ordering must never select an operation or schema, and `toolContract` is the named accessor that enforces it. * test(v8): cover the new tool in the absolute-path boundary test Main added `tests/boundary/mcp/absoluteLocalPathTools.test.ts` after this branch was opened. It enumerates every tool whose MCP `tools/list` schema must reject a relative path, and `inspect_v8_code_cache` uses `absoluteLocalPathStringSchema`, so the tool belongs in that list. Adding it keeps the enumerated field set complete instead of leaving one absolute-path tool untested. `tests/boundary/mcp` + v8 + contracts: 256 passing. `tsc` 0; `build` 0; `knip` 0; `docs:check` 0; `npm run lint` 0 errors. * perf(v8): read the payload by index instead of allocating per byte `for (const [index, byte] of payload.entries())` allocates a two-element array for every byte of the cache payload. On a 1 MiB payload that allocation dominated the scan. Measured with a real cache built from this parser, alternated so the two forms run against the same input: | payload | entries() | indexed reads | | --- | --- | --- | | 1 MiB | 15.5ms | 4.4ms | That is 3.49x on the scan, and the gap widens on smaller payloads where the allocation cost is a larger share (7.3x at 64 KiB in a synthetic run of the same shape). The scan is the whole cost of `inspect_v8_code_cache` once the file is read, so this is the tool's own hot loop. Output is unchanged: the parser returns one retained string for the 64 KiB fixture, and `src/v8` plus the CLI and MCP boundary files stay at 21 passing. `tsc` 0; `build` 0; `npm run lint` 0 errors. * fix(v8): admit only decimal tokens on the numeric CLI options Main added `tests/boundary/cli/cliDecimalArguments.test.ts` after this branch was opened. It enumerates every numeric CLI option and requires `0x10` and `1e3` to be rejected, because plain `Number` coercion would read them as 16 and 1000. This command's three options forwarded the domain fields unchanged, so the test failed on them: inspect-v8-code-cache options.maxFileBytes: Parser.parse accepts 1e3 inspect-v8-code-cache options.maxStrings: Parser.parse accepts 0x10 inspect-v8-code-cache options.maxStrings: Parser.parse accepts 1e3 inspect-v8-code-cache options.minStringBytes: Parser.parse accepts 0x10 `cliDecimalField` publishes the input view that parses the decimal token first and then lets the domain field own bounds, defaults and help, which is what the other numeric commands use. `tests/boundary/cli/cliDecimalArguments.test.ts` now passes. `tsc` 0; `build` 0; `npm run lint` 0 errors; `knip` 0; `docs:check` 0. * fix(v8): recognize module caches, and advise from remaining headroom Two review findings, each measured on Node 24.22.1 before changing anything. **A module cache was reported as a script.** `vm.SourceTextModule.createCachedData()` sets bit 30 of the source-length word, which the source-length mask `0x1fffffff` leaves out, and the producer check only tested the function bit at bit 29: | producer | source word | | --- | --- | | `vm.Script` | `0x00000013` for a 19-unit source | | `vm.compileFunction` | `0x20000009` for a 9-unit source | | `vm.SourceTextModule` | `0x40000013` for a 19-unit source | So a real module cache was labelled `script`, and `vm.Script` rejects that container. The producer enum gains `module` and the module bit is tested first; the source length is unchanged because both flags sit outside the mask. **The budget advice recommended a retry that could not work.** Both limits have ceilings — `max_strings` at 65536 and `min_string_bytes` at 256 — so an admitted file with more than 65536 runs of 256 bytes reached the branch while every suggested retry was a rejected input. `observationBudgetRemediation` now reads the remaining headroom and only recommends a change that is still admissible, falling back to selecting a smaller artifact when neither can move. The parser message states the fact and leaves recovery to the provider, which is the only place that can see both budgets. `MAX_INSPECT_V8_CODE_CACHE_MIN_STRING_BYTES` names the `min_string_bytes` ceiling that the schema previously spelled inline, so the advice and the schema cannot drift. `docs/v8-code-cache.md` describes both flag bits. `src/v8` plus the CLI and MCP boundary files: 28 passing; 7 new cases fail on the base commit. `tsc` 0; `build` 0; `npm run lint` 0 errors; `knip` 0; `docs:check` 0. * refactor(v8): decide the budget recovery as data, and test the decision The remediation tests matched fragments of the advice sentence and banned old phrases. `docs/testing.md` asks for error codes, recorded limits and recovery behaviour instead of frozen prose, so a harmless reword could fail those tests while wrong advice passed them. `observationBudgetRecovery` now returns the decision — which retries remain within the schema's admitted range, and whether a smaller artifact is the only option — and `observationBudgetRemediation` renders one sentence from it. The tests assert the decision, so the wording is free to change and an incorrect choice cannot pass. This also fixes a bug in the previous shape: it took the `min_string_bytes` branch to mean "lower it", but that branch is reachable with the value already at the schema minimum, where lowering is not a legal retry. The decision now carries `canLowerMinStringBytes` and a `requiresSmallerArtifact` flag for that case. 5 cases, all failing on the base commit, covering both headroom states, either ceiling alone, and the run-length floor. `src/v8` plus the CLI and MCP boundary files: 28 passing. `tsc` 0; `build` 0; `npm run lint` 0 errors; `knip` 0; `docs:check` 0. * ci: rerun after the wakaru release download recovered The only failing job on this branch was `javascript_recovery / real-javascript-recovery`, which failed on `curl: (22) The requested URL returned error: 504` while downloading `wakaru-linux-x64.tar.gz`. The release is reachable again (HTTP 200), so this empty commit re-triggers the lane. * docs(v8): place the guide in the tree and record the feature The guide sat at `docs/v8-code-cache.md`, outside the `docs/guides/` tree every other guide uses, and neither index linked it, so it could not be reached from `docs/README.md` or the VitePress sidebar. Moved to `docs/guides/javascript/`, linked from both indexes, and recorded the feature under `docs/specs/v8-code-cache-inspection/` with `spec.md`, `plan.md` and `tasks.md`. The spec keeps only the premises that were measured on Node 24.22.1 — the three producer words and the retained-string layout — rather than restating the implementation. `npm run docs:check` 0; `npm run format:check` 0; `tsc` 0; `build` 0; `lint` 0 errors; `knip` 0. V8 reader, CLI and MCP files: 28 passing. * docs(v8): use the documented spec path and test the advice against its struct Two review findings. **The records were at `docs/specs/v8-code-cache-inspection/`.** `AGENTS.md:73` asks for `docs/specs/<domain>/<feature-slug>/`, so they move to `docs/specs/v8/code-cache-inspection/`. **The verification section claimed more than the tests do.** It said the producer cases were built from real `vm.Script`, `vm.compileFunction` and `vm.SourceTextModule` output, and that the budget advice was compared with its owning formatter. In fact only `vm.Script` and `vm.compileFunction` are real producers — the module case sets the measured bit on a `vm.Script` cache — and only `observationBudgetRecovery` was asserted. Corrected the claim, and closed the second gap rather than only describing it: `observationBudgetRemediation` now has four cases holding the sentence to the struct it renders from. The sentence must not advise a knob the struct closed, and must mention a smaller artifact exactly when the struct requires one. Verified by making the sentence advise `max_strings` after the ceiling was reached — the test fails. Wording stays free; wrong advice does not. `npm run docs:check` 0; `npm run format:check` 0; `tsc` 0; `build` 0; `lint` 0 errors; `knip` 0. V8 reader, CLI, MCP, absolute-path and decimal-argument files: 34 passing. * fix(v8): restore the absolute-path contract entry after the rebase The rebase onto `main` conflicted in `tests/boundary/mcp/absoluteLocalPathTools.test.ts`: `main` gained `inspect_api_contract`, `inspect_devicetree`, `inspect_playwright_trace` and `inspect_python_pyc` in both the contract list and the path-field map while this branch added `inspect_v8_code_cache`. Both sides are additive, so both are kept. Without the resolution the branch carried conflict markers, which made `mergeable_state` `dirty` and every check on it meaningless. Re-verified after the rebase: `tsc` 0, `build` 0, `format:check` 0, `lint` 0 errors, `knip` 0, `docs:check` 0, and the V8 reader plus its CLI, MCP, absolute-path and decimal-argument boundary files at 34 passing. * test(v8): pin the producer from the runtime's own header word An independent reviewer ran these files on Node 22.23.4, 24.22.1 and 26.12.1. Only 22.23.4 failed, on `separates the compileFunction producer bit from the source length`, expecting `compile_function` and reading `script`. The cause is the fixture, not the parser: `vm.compileFunction(...).cachedData` carries `0x20000009` in the source-length word on Node 24.22.1, with bit 29 set, while Node 22.23.4 does not set that bit at all. Whether a given V8 build marks the word is a build detail, so asserting one build's spelling made the check version-dependent. The case now reads the word the runtime produced and asserts the parser maps it correctly, and keeps the part that must hold everywhere: the source length is the low bits, so it equals both `word & 0x1fffffff` and the source's length. The flag meanings themselves stay pinned by the synthetic `withHeaderWord` cases above, which do not depend on any build. Reverting the parser's `compile_function` arm fails this case and `keeps the function and module flags distinct`. `src/v8` plus the CLI, MCP and absolute-path boundary files: 24 and 34 passing. `tsc` 0; `build` 0; `npm run format:check` 0; `npm run lint` 0 errors; `knip` 0. * fix(v8): report an unflagged producer word as unknown, not as a script `vm.compileFunction` marks bit 29 of the source-length word on Node 24.22.1 (`0x20000004` for a 4-unit source) and does not mark it on Node 22.23.4 (`0x00000004` for the same call). Whether a V8 build marks a function cache is therefore a build detail, and an unflagged word is either a `vm.Script` cache or an unmarked function cache. The container carries nothing that separates the two. The reader reported `script` for that case, which is a confident claim the bytes do not support. My previous commit made the test accept whichever value the runtime produced, which recorded the ambiguity instead of resolving it in the contract. `unknown` is now a producer value, the unflagged word maps to it, and the observed word stays in `header.fields` for a caller that trusts a specific build fingerprint. A limitation line says the same in the report itself. `header.producer` widens to `["script", "compile_function", "module", "unknown"]`; `script` is no longer produced, because no unflagged word can be shown to be one. The spec's producer rule and its verification notes are updated to match. Reverting the mapping to `script` fails 2 of the 24 `src/v8` cases. `src/v8` with the CLI, MCP and absolute-path boundary files: 33 passing. `tsc` 0; `build` 0; `docs:check` 0; `npm run format:check` 0; `npm run lint` 0 errors. `knip` reports one unused export (`builtinDataEffectMethod`) on `main` as well, so it is not from this change. * fix(artifacts): retain completed identity across native parser failures * docs(v8): preserve unknown unflagged cache provenance --------- Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
|
ab1b9e35d0 |
fix(entropy)!: own selected Windows path spelling at admission (#9393)
* fix(evidence)!: keep one Evidence identity across Windows path case On Windows, an existing caller path now takes its filesystem spelling when that spelling differs from the resolved path by letter case alone. One file spelled `C:\Temp\x.bin`, `c:\temp\x.bin` or `C:\TEMP\X.BIN` produces one `parameters.path`, `normalized_result` path and `evidence_id`, matching the existing separator and `.` folding. `caller_input_path` keeps the input. Links, junctions, `subst` drives, 8.3 short names and paths that do not exist yet keep the caller spelling, so no link is followed and requested output names keep their case. BREAKING CHANGE: Windows Evidence produced from a path spelled in a case other than its filesystem spelling now gets a different `evidence_id`. Fixes #9373 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * fix(entropy): keep Windows casing lookup optional * ci(entropy): select Windows identity verification --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
morluto
|
bf271ba3fb |
fix(capture): retain safe caller path selections (#9626)
* fix(cli)!: retain repeated WASM input paths BREAKING CHANGE: caller_input_paths values may now be ordered arrays of strings as well as strings. Clients must accept both forms. * fix(capture): retain safe caller path selections * fix(evidence): retain auxiliary selections in direct summaries * test(wasm): separate SDK schema and acquired-input coverage |
||
|
Taksh Kothariandmorluto
|
425581e838 |
fix(cli): validate paths before capture file admission (#9318)
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
|
908d9c0726 |
fix(javascript): keep export bindings out of module views (#9486)
* fix(javascript): keep export bindings out of module views Exported symbol bindings reuse the javascript-module node kind, so the modules collection counted one row per export alongside the owning module. On a CommonJS or ESM file with N exports, inspect_analysis_view reported N+1 module units, module pages listed indistinguishable rows sharing the module path, and item selection by selector.path failed with "Multiple modules have path". Treat a node whose observations carry semantic_role export-binding as an export of its owning module, not a module unit, and exclude it from module entries. Export bindings remain distinct graph entities for export-shape comparison and version matching. Validation: a domain regression in analysisView.test.ts and a CLI acceptance regression over a real CommonJS analysis both fail on main and pass with this change. Also ran the analysis view, MCP boundary, CLI boundary and JavaScript acceptance suites (125 tests), npm run check:fast, and the MCP catalog schema test after regenerating .cache/mcp-tool-catalog.json. * test(javascript): run the export-binding view case on the shared CLI The case starts a compiled CLI process for each of its three commands, paying startup and tool registration three times per run. The fixture already provides a file-scoped process for calls with no stdin or explicit timeout, so use it. * style(javascript): format the constructed-evidence type in the module-view case The acceptance case's inline graph type did not match what oxfmt emits for the 80-column limit, so `format:check` refused the tree. Reflow only; the assertions are unchanged. * style(cli): wrap the native data type path description The `inspect-native-data-type` argument description line is 82 columns, over the repository's 80-column `.oxfmtrc.json` limit, so `npm run format:check` fails on main and on every branch merged from it. It came in with #9465 and the following commits to this file did not touch it. Reformat the one call the formatter objects to; no behaviour change. --------- Co-authored-by: 2SinceJuly <202121119+2SinceJuly@users.noreply.github.com> Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
morlutoandhhang
|
6a24e801ea |
fix(elf): qualify Linux CORE notes by the declared ABI (#9613)
* fix(elf): qualify Linux CORE notes by the declared ABI * test(mcp): cover source fitting and omitted delivery budgets --------- Co-authored-by: hhang <169027046+xu91102@users.noreply.github.com> |
||
|
|
7baed55130 |
fix(doctor): probe providers with the selected client's registration environment (#9108)
* fix(doctor): probe providers with the selected client's registration environment rea doctor read provider settings only from the terminal running it. A client registration that carries GHIDRA_INSTALL_DIR or REA_JADX_MCP_CONFIG under its own env block therefore looked unconfigured: doctor answered not_configured while the server that client starts inherited the settings and worked. This is the recovery command the docs point agents at, so the false negative sent them back to re-apply fixes that were already in place. When the caller names exactly one client, read that registration's effective environment and hand it to the provider checks. Terminal variables that providers resolve outside server config (PATH, SYSTEMROOT) stay inherited, and running doctor without --client still reports this terminal. A registration the loader cannot read keeps its own registration check as the answer instead of being re-reported as four missing providers. Fixes #9102 * fix(doctor): keep setup's terminal settings and a failed close on the host Greptile found two regressions in the selected-client environment read. Provider probes now start from the terminal and overlay the selected registration, instead of replacing the terminal with the registration alone. `rea setup --client` passes its own client selection through the same doctor call, so on a first run with GHIDRA_INSTALL_DIR exported and no registration yet, the replacements made setup report providerConfigured: false and offer Hopper over the selected BYO provider. The registration still wins for every value it records. The registration read now runs inside the caller's resource scope and, when the close of the configuration file fails, retains the returned owner there so host.close() retries it. Dropping that owner left an unretired handle with no way to report the unfinished cleanup. * fix(doctor): leave a registration's expression settings unknown Greptile found the previous fallback filling a name the selected client records as a Cordis expression with the terminal's value. `dsh` keeps such a value as `!!js ...`, which is not a literal; the parser reports the name in `unresolvedSettings`, and merging the terminal into it then made probes certify a setting the client owns and this process cannot read. Unresolved names are now removed after the overlay, and an unreadable, malformed, or unresolvable registration returns no environment at all, so the caller keeps the terminal instead of a partly-known one. Provider checks report those settings as unconfigured, which is the honest answer for a chain that resolves outside REA. dsh's missing registration still inherits the terminal, which is the case `rea setup` depends on. * fix(doctor): fold registration settings by platform identity `selectedProviderSettings` overlaid a selected client's registration on the terminal with an object spread. On Windows a registration can record `ghidra_install_dir` while the terminal exports `GHIDRA_INSTALL_DIR`; the spread kept both spellings, and since provider checks read the canonical uppercase names, doctor either certified the terminal's install or reported the setting as unconfigured. An expression the registration marks unresolved was skipped under one spelling while its terminal value survived under the other. Merge the two layers through `mergeEnvironmentOverrides` and fold the result with `snapshotEnvironment`, so the registration's value wins regardless of how it spells the name and the provider sees the names it reads. Fold the unresolved names the same way before deleting them, and export the helper so the two platform identities can be covered directly. * style(doctor): wrap the long import in the runtime test oxfmt wants the three imported names on their own lines; the previous commit left the single-line form, which fails `npm run format:check`. * test(codebuddy): construct the registration command like its sibling clients `npxRegistrationCommand` was removed in |
||
|
Ziqian zhangandmorluto
|
94728785ea |
fix(setup): refuse multi-document DSH patches (#7512)
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com> |
||
|
gongfpp
|
4df630e16a |
fix(javascript): keep unisolated program-var route values unknown (#6778)
* fix(javascript): mark unisolated program-var route values unproven `JavaScriptRouteScan.#resolve` now records the identifier whenever a value proof depends on a program-scope `var` in a script without a proven isolated runtime. That binding is a global-object property, so `globalThis.NAME` or an alias can replace it after the source initializer. The mark lives in the shared resolver, so every consumer degrades from one fact instead of each one opting in: - path segments and navigation destinations stop asserting the source value and report an unknown interpretation; - an index flag stays unproven, as before; - a route container keeps its source route set and reports an incomplete interpretation, so `var routes=[...]` no longer loses every route. The source initializer is still followed, which is what lets a container consumer keep its facts while disclosing the gap. * fix(javascript): keep literal child routes behind an unisolated children array Rejecting the whole children configuration marked its records as claimed by the interpretation, so the automatic-root loop skipped them and the literal child paths disappeared even though their absolute paths need no parent proof. Keep the source records reachable and let the walker report the unproven ancestry as an interpretation gap. * test(javascript): bound the index-route CLI acceptance matrix The flag matrix runs sixteen CLI invocations in one test and passed at 20.7 s against a 30 s default timeout, so it timed out on a loaded runner. The same matrix is already asserted per case in javascriptRouteScan.test.ts, where each case is an in-process parse; this file keeps one resolved case and one unresolved case to prove both branches of the public CLI view. * fix(javascript): retain ownership of unproven route literals Keep source ownership separate from runtime value proof so rejected program-var route segments cannot reappear through the generic constant pass. Cover path, route, typed and alias forms, retain navigation controls and independent literals, and verify inventory and public CLI projections. Run the existing CLI index cases independently so repeated process startup does not consume one combined test deadline. Preserve both cases and all assertions. * fix(javascript): keep unisolated navigation data out of route literals An unisolated program-var slot holds data, not a route. Refusing its runtime identity released the object back to the independent route scan, which then reported a query payload as a nested route. Report the object as consumed navigation data and stop following it as a path. * fix(javascript): follow rejected navigation data without trusting it Claiming the whole object for a rejected value also claimed its sibling properties, so an independent route in the same object disappeared. Keep the property selection: ownership follows the slot, and only the route consumers refuse an unproven runtime identity. |
||
|
morluto
|
4a1c4a59ba | test(discovery): separate recovery guidance and owned home state | ||
|
morluto
|
5d5fe72218 | fix(plist): reserve Evidence framing in nesting allowance | ||
|
morluto
|
bd0642e0a4 | test(javascript): give loader fixtures explicit source context | ||
|
morluto
|
99d00d1041 | perf(tests): reuse CLI workers for Rust and certificate cases | ||
|
morlutoandShuaixinHuang
|
b46f4fcebd |
fix(javascript): preserve loader authority for exact suffixes (#9477)
* fix(javascript): require exact loader suffix case * fix(javascript): share exact loader mode across import resolution * test(process): align symlink fixture with filesystem checkpoints * test(javascript): normalize the module resolution oracle path separators * test(frida): assert the CLI spawn Evidence contract * test(mcp): use an unsupported analysis-view parent * test(browser): separate source-location line bounds from parity * fix(javascript): preserve uncertain loader provenance * test(javascript): cover uncertain require resolution * fix(javascript): retain conditional Electron evidence * fix(javascript): preserve timed semantic provenance * fix(javascript): retain conditional builtin effects * test(javascript): distinguish bundled host authority * fix(javascript): preserve nested loader invocation identity * test(javascript): qualify Electron fixture authority * fix(javascript): preserve loader invocation authority * fix(javascript): respect source order for resolver escapes * fix(javascript): check nested loader callable mutations * fix(javascript): retain conditional constructor requests * fix(reference): retain constructor loader relationships * fix(reference): follow nested loader callable paths * refactor(javascript): share callable mutation interpretation * fix(javascript): require exact package loader suffixes * refactor(javascript): unify conditional relation evidence * refactor(javascript): unify conditional node evidence * fix(javascript): retain unknown authority for mixed case sources * fix(javascript): preserve loader uncertainty for dotfiles * fix(reference): classify javascript suffix dotfiles * test(reference): cover suffix dotfiles through cli * fix(javascript): retain equivalent conditional built-in origins * fix(javascript): preserve wrapped resolver operation * fix(javascript): retain nested conditional built-in receivers * test(javascript): declare CommonJS in native addon fixtures * fix(javascript): retain wrapped resolver candidates * style: format merged native command schema * fix(javascript): preserve uncertain loader authority * test(javascript): require candidate preload evidence * test(javascript): cover reference loader member writes --------- Co-authored-by: ShuaixinHuang <1836656514@qq.com> |
||
|
morlutoandFelix
|
52fe342c73 |
fix(evidence): preserve subjects in derived workflows (#9492)
* fix(javascript): preserve verified trace evidence subjects Co-authored-by: Felix <3937061056@qq.com> * fix(evidence): preserve selected subject names in views * fix(evidence): preserve subjects across related workflows * fix(evidence): retain inventory projection subjects * fix(managed): validate projected graph root identity * style: format evidence and native command code * fix(javascript): validate static graph root identity * test(javascript): update feature trace graph expectations --------- Co-authored-by: Felix <3937061056@qq.com> |
||
|
morluto
|
0b42dc56d1 | perf(tests): reuse CLI sessions for analysis data cases | ||
|
morluto
|
a961db280f | fix(filesystem): explain refused file symlinks (#9476) | ||
|
Ryan S
|
94182f331e |
feat(go): recover bounded Go string literals from (ptr, len) headers (#9206)
* feat(go): recover bounded Go string literals from (ptr, len) headers * fix(go): deduplicate header scan spans, bound rejected cache, and reject Cf * fix(go): allow U+200C ZWNJ and U+200D ZWJ in Go string literal recovery * fix(go): index Mach-O load regions, validate MCP schema, and add feature specs * test(go): cover Mach-O region-index cancellation and load-region boundaries |
||
|
morluto
|
5553acfe50 | fix(elf): report incomplete kernel module coverage (#9475) | ||
|
ShuaixinHuang
|
90c95a6b64 |
fix(apk): validate the Zip64 end record size (#9432)
* fix(apk): validate the Zip64 end record size * test(ghidra): assert current provider limitations |
||
|
morluto
|
25d3c7237f | fix(paths): enforce selected path contracts at admission (#9465) | ||
|
morluto
|
5044e20171 |
test: focus diagnostics and evidence checks on behavior (#9303)
* test: focus configuration diagnostics on their owners Preserve shape errors, selected paths, recovery guidance, and unchanged-file behavior across parser, setup, and uninstall boundaries. * test: focus artifact and evidence checks on behavior Retain format, identity, malformed-input, and cleanup coverage while removing fixed prose and closing the admitted test source. * refactor(debugger): use configured session admission limits * test: focus evidence and diagnostics on behavior * fix(test): restore verifier export and formatting * test(config): verify debugger limit ranges * test(rizin): keep fixture executable authoritative * test(rizin): isolate debugger fixture environment |
||
|
morlutoandCursor Agent
|
b81da90354 |
feat(native): page retained procedure references (#8588)
* feat(native): page retained feature traces * fix(snapshot): retain linked source evidence for replay * test(snapshot): verify linked foreign evidence stays scoped * feat(native): page retained procedure references * docs(mcp): clarify imported residual questions * docs(mcp): guide trace reference classification * fix(cli): report result sources when Evidence publication fails A trace keeps linked sources on the result. Use those records for the undelivered source closure as well as the bundle, so a refused output does not report them as not captured. * ci: rerun pull request checks The merge of origin/main did not start the CI workflow for this head. * fix(native): document shared procedure endpoints and the required selector Describe kind shared and items.endpoints instead of inline procedure copies. State that procedure-references requires a positional selector or --procedure. * fix(native): count procedure-reference endpoints before keeping a row Include the shared endpoint table in the page budget, with JSON punctuation, and store a new procedure only when its row is accepted. * fix(native): measure a procedure endpoint once before its comparison key Size the candidate before building one canonical identity, and reuse that key for the selected-procedure and shared-endpoint checks. * fix(native): reject an oversized procedure endpoint before canonicalizing it When the JSON size already exceeds every stored identity, skip the canonical string and keep the row out of the page. * fix(native): describe procedure-reference direction and coverage Avoid execution wording in the tool description, and give the CLI direction option the help text the catalog requires. * fix(native): compare endpoint size with UTF-8 identity length Keep one running maximum of stored identity bytes. A multibyte selected procedure is recognized when its JSON no longer fits the page. * test(cli): count procedure-references as a dedicated command The MCP tool has a CLI command, so it is no longer an MCP-only exception. * test(native): catch the string-length endpoint comparison Use a 200,000-character multibyte selected procedure and a shorter ASCII endpoint so the old JavaScript length check drops the selected row. * ci: rerun after the binary-layout toolchain timeout The required check failed because the toolchain install exceeded 15 minutes. The same job completed on the previous head, and this commit does not change the procedure-reference behavior. * test: stabilize capture decoder exhaustion coverage --------- Co-authored-by: Cursor Agent <cursoragent@cursor.com> |
||
|
morluto
|
352a846127 |
test: reduce brittle assertions and test-only runtime paths (#9289)
* test: focus assertions on behavioral evidence Prune redundant and wording-coupled checks while retaining diagnostic, parser, recovery, and cleanup coverage. * refactor: remove test-only wrappers and duplicate checks Exercise production operations through test fixtures, remove unused adapters, and keep managed-member uncertainty guidance under its existing owner. * test: focus process validation on diagnostic facts * refactor: remove test-only runtime helpers Move semantic evidence fixture setup out of runtime and remove unused Windows settle recovery helpers while checking configured defaults independently. * test: focus artifact diagnostics on format facts * test: verify retained cleanup guidance delivery |
||
|
morlutoandCursor Agent
|
31278c7b5a |
fix!: qualify call paths by provider discovery (#8443)
* fix!: qualify call paths by provider discovery * fix: preserve compact call coverage schema guidance * fix: preserve call discovery reasons in path evidence * fix(evidence): retain discovery limits in call verification * test: declare discovery in reconstruction controls * fix: preserve call discovery across facet limits --------- Co-authored-by: Cursor Agent <cursoragent@cursor.com> |
||
|
morlutoandCursor Agent
|
60f9a4c69a |
fix(native): show function match limitations by default (#8558)
Co-authored-by: Cursor Agent <cursoragent@cursor.com> |
||
|
morluto
|
1596d6a275 | test: reduce wording coupling and prune redundant coverage | ||
|
morluto
|
80427af8cc | test: repair snapshot and setup isolation checks | ||
|
morlutoandCursor Agent
|
2e2f981ccf |
test: assert absolute-path facts without frozen sentences (#9231)
* test: assert absolute-path facts without frozen sentences Sibling path tests now require the field path, invalid_format, and regex expectation, and they record that each negative case ran. NUL rejection follows the argument that contains the byte instead of searching for the word NUL. * test: retain actionable NUL path diagnostics --------- Co-authored-by: Cursor Agent <cursoragent@cursor.com> |
||
|
morlutoandCursor Agent
|
562779bd71 |
fix(mcp): advertise absolute paths for tools that reject relative inputs (#9179)
* fix(mcp): advertise absolute paths for tools that reject relative inputs The domain schemas accepted relative paths, then handlers rejected them with isAbsolute. Those inputs now use absoluteLocalPathStringSchema, and the duplicate handler checks are gone. * fix(javascript): own recovery path admission in the schema Relative recovery paths now fail in absoluteLocalPathStringSchema, including the additional_paths index. Drop the duplicate handler check, and make the CLI, MCP, and tools/list tests assert that schema fact. * test: assert absolute-path facts instead of error prose Recovery and web-script tests keep the field, format code, and regex expectation. They no longer freeze the schema's explanatory sentence. --------- Co-authored-by: Cursor Agent <cursoragent@cursor.com> |
||
|
morluto
|
99db80a13e | fix(cli): reject non-decimal integer tokens on every numeric option (#9178) | ||
|
ShuaixinHuang
|
0fc35a098a |
test(javascript): normalize module resolution oracle paths (#9159)
* test(javascript): normalize module resolution oracle paths * test(process): align symlink fixture with filesystem checkpoints * test(frida): assert the CLI spawn Evidence contract * test(mcp): use an unsupported analysis-view parent * test(browser): separate source-location line bounds from parity |
||
|
gensokyo-devandgensokyo-dev
|
98c589e420 |
perf(tests): reuse bounded CLI workers for native matching errors (#9163)
Co-authored-by: gensokyo-dev <gensokyo-dev@users.noreply.github.com> |
||
|
gensokyo-dev
|
bacc1e3846 |
perf(mcp)!: defer retained views in progressive catalogs (#9151)
Keep full schemas and canonical analysis unchanged while loading the cross-domain view through existing named discovery. Qualify real recent-tool workflows with token ledgers, latency, CPU, family RSS and retained-follow-up controls. BREAKING CHANGE: The opt-in hot-cold catalog no longer preloads inspect_analysis_view. Discover it before calling it, batch with its producer when needed, or use the unchanged default full profile. |
||
|
gensokyo-dev
|
1480a7c72e | docs(firmware): qualify component comparison workflows (#9143) | ||
|
RRFZ
|
55ee36809d |
fix(javascript): apply package type scope to application sources (#7883)
* fix(javascript): apply package type scope to application sources Application analysis parsed .js sources without their package scope, so an unbound require in a type-module package became a resolved CommonJS dependency. The parent now observes each source's nearest same-container package.json, stopping at node_modules, and the worker parses with that package type as reference imports already do. Fixes #7323 * fix(javascript): keep package type unknown behind unavailable manifests The package-type lookup indexed only readable files, so a nearer package.json whose bytes were unavailable was skipped and an ancestor's "type" applied to sources it does not own. Index unavailable manifests by container and directory too, and stop the walk with "unknown" when one is the nearest owner, as the Next app-context lookup already does. * fix(javascript): keep Electron preloads outside module package scope Electron's preload loader ignores a package's "type": "module" and gives a .js preload `require`; ESM preloads need the .mjs suffix. Applying the package type to every .js source marked such a preload as an ES module, so its `require("electron")` edges and IPC pairing disappeared. A source that requires an Electron module now keeps the context REA inferred before package scope was applied. Node-loaded sources in a module scope still drop unbound `require`. * fix(javascript): admit Electron script loaders only on renderer evidence The previous commit dropped a module package type whenever a source required an Electron module. A Node-loaded ES module that requires Electron, or one whose Electron call goes through a local `require`, then regained the false require edges the package scope removes. The parser applies the package type unconditionally again. The worker re-reads a package-module script that requires an Electron module once outside the package scope. The source keeps that context only when the static pass there directly observes a contextBridge exposure or a renderer IPC operation, which only Electron's renderer-side loaders provide. * fix(javascript): check Electron renderer evidence without a second static pass |
||
|
2SinceJulyand2SinceJuly
|
6c08320654 |
feat(process): add a non-executing dry run to capture_process_scenario (#8469)
* feat(process): add a non-executing dry run to capture_process_scenario Preparation prompts asked an agent to prepare a scenario without running it but could only suggest the tool that launches the target. Add dry_run: true so the CLI and MCP resolve the executable, working directory, observation paths and admission checks, then return planned Evidence under the rea.process-scenario-plan predicate. Planned Evidence is not capture Evidence and comparison tools refuse it as a capture. The preparation prompt now suggests the dry-run call, and the scenario identity commitments exclude the operation-only field. Fixes #8398 * fix(process): redact planned sensitive input and share admission Address review findings on the dry-run contract: - Redact sensitive input events in planned Evidence and preserve only their byte length. - Recheck cancellation after the last awaited admission step. - Run the sensitive-matcher and filesystem-root admission checks that a capture performs before launch. - Assert the prompt advertises the dry_run field without freezing explanatory wording. * test(process): hold the final admission step for cancellation Add a preparation host seam so the regression can cancel while the last filesystem admission call is pending, then assert the returned AnalysisCancelledError retains the caller's reason. * refactor(process): remove the unused plan alias Keep the plan schema as the single exported contract and satisfy the unused-export check. * fix(process): report admission-time aborts as planning cancellation `prepareProcessScenarioPlan` returned the admission error verbatim, so an abort that landed while `admitFilesystemObservationRoots` awaited its `lstat` surfaced as a cancelled `ProcessCaptureError` with `operation: process_capture` even though no process ran. Convert that result through the same `cancelled()` projection the neighbouring checks use, and pin it with a test. * fix(contracts): keep the process result invariants on the dry-run union The output schema for `capture_process_scenario` became a union of the capture and plan shapes, and the union no longer carried the description that advertised the cross-field invariants to MCP clients — the boundary test reads `normalized_result.description`. Attach a description to the union that names both variants and keeps the capture invariants verbatim. * test(process): exercise the post-admission cancellation check The previous regression aborted immediately after starting preparation and released the held admission afterwards. Preparation awaits runtime path resolution first, so that abort was always observed by the pre-admission check; deleting the post-admission "if (aborted()) return cancelled();" still left the test green. Signal from the admitRoots seam when admission is actually entered, abort only after that point, then let admission resolve ok. Preparation now has to reach the check between a successful admission and the ok result, and the assertion pins both the AnalysisCancelledError tag and the caller's reason as its cause. Verified red-green: removing only the post-admission check fails this test (expected true to be false at prepared.ok) while the other two pass, and restoring it passes all three. * test(process): gate the admission seam on acknowledgement Race the admission acknowledgement against planning's own settlement and a bounded deadline, and release and join planning in `finally`, so a planning result that arrives before admission fails the test instead of stalling it until the suite timeout. * fix(contracts): hold plans to the scenario refinements and redact inputs `processScenarioPlanSchema` described the resolved dry-run response but dropped every cross-field check the input contract applies, so a plan could advertise out-of-order events, an event past `timeout_ms`, or a terminal beyond the renderer cell budget while claiming it had been validated. Extract those checks into `scenarioRefinements` and refine the plan's scenario with them. The plan's input event also carried `sensitive` and an optional `input_bytes`, which accepted a sensitive event that still shipped its data. Model the public and sensitive forms as separate exact variants: public input forbids `input_bytes`, sensitive input requires an empty `data` plus its byte length. --------- Co-authored-by: 2SinceJuly <202121119+2SinceJuly@users.noreply.github.com> |
||
|
gensokyo-dev
|
5303199163 |
feat(api-contract): inspect local GraphQL SDL and field references (#9140)
* feat(api-contract): inspect local GraphQL SDL and field references * fix(api-contract): retain caller metadata and bound selection work |
||
|
gensokyo-dev
|
985b9eae96 |
feat(browser): import saved Playwright v10 trace evidence (#9127)
* feat(browser): import saved Playwright v10 trace evidence * fix(browser): preserve original Playwright ZIP member names |
||
|
gensokyo-devandgensokyo-dev
|
afc9cfe9ff |
perf(tests): reuse bounded CLI workers in mobile archive checks (#9116)
Co-authored-by: gensokyo-dev <gensokyo-dev@users.noreply.github.com> |
||
|
gensokyo-dev
|
91d73a113e |
feat(contracts): inventory offline OpenAPI operation declarations (#9104)
* feat(contracts): inventory offline OpenAPI operation declarations * docs(contracts): link fixture provenance outside the docs site * fix(contracts): describe the shared API document byte allowance * fix(api-contract): preserve safe diagnostics and completed source facts |