rea-agents 6.5.0 passed an absolute caller path to Windows native
snapshot admission verbatim, so a spelling with forward slashes or
repeated separators reached the native component walker, which rejects
empty components as "Ambiguous Windows path component is unsupported"
(Win32 50). Current main already resolves every Windows caller path with
win32.resolve (fc791bb60), which collapses those spellings.
Pin that boundary: absoluteCallerPath normalizes four separator spellings
of one absolute Windows path, and on Windows parseBinaryTarget returns the
same canonical sourcePath for each of them. The Windows test fails when
the 6.5.0 verbatim-absolute-path rule is restored.
Fixes#4707
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Reuse stable-file hashing and exact cleanup leases for target verification, join admitted call workflows before tracer close, and block replacement work while descriptor cleanup remains unresolved. Preserve physical failures, exact cancellation reasons and completed call facts; remove the duplicate reader and AbortError-name heuristic.
Validate real inert-file cleanup controls and controlled tracers through provider and MCP boundaries. No LLDB execution, real OS close-failure or Windows qualification.
Refs #4740
Reuse test-owned stdio MCP connections within two stateless source-analysis matrices while retaining every original CLI/MCP scenario, schema/Evidence checks and test-finished cleanup. Add cross-input identity controls.
Validation: 23 focused public-workflow tests and complete check:ci passed; stale-response fault detected. Exact-head hosted CI and available review passed. Pinned workload measurements and their limits are recorded in the PR.
Refs #4737.
Bind signature inspection to its named domain result and retain completed facts in typed cancellation. Check cancellation after descriptor and root finalization, preserve exact caller reasons, and keep physical failures primary. Use explicit cancellation checkpoints instead of signal-state guesses.
Validate root and descriptor cleanup, retries, late root handoff, opaque reasons, provider projections, and existing real codesign/CLI controls. No real OS close-failure or Windows qualification.
Refs #4735
Use shared owned file operations for plist admission and source probes. Join admitted work during close, retain exact descriptor leases through retries, and preserve physical failures, cancellation reasons and completed converter output.
Validate with controlled cleanup faults and native plist CLI workflows on inert data. Real OS close-failure incidence and Windows behavior remain unverified.
Refs #4723
Own decoded node columns, attributes and parent validity once for web, Electron and SVG interpretation. Reject dangling, self and cyclic parent facts; preserve source positions and independently unknown fields. Remove fallback conversions and duplicated string/attribute interpretation.
Validation: baseline public CLI controls failed for four malformed cases with valid control passing; final 103 focused checks, 5 final built CLI controls, 1 late-detach cancellation control (20 skipped), check:fast and docs:check passed. Fake CDP producers and generated local files; no real browser/Electron or Windows qualification.
Refs #4709
* fix(native): report unidentified text plists as unknown
Require explicit XML identification in file output instead of treating all
text as XML. Preserve decoded plist values and source-format provenance
when plutil falls back to XML conversion.
Add provider-boundary regressions and real macOS CLI controls for OpenStep,
data-bearing OpenStep, XML, UTF-16 XML and binary property lists.
Fixes#4294
* fix(native): keep XML plists without a declaration labeled xml
`file -b` names XML only when a document declaration is present, so the
narrowed `\bXML\b` check reported a declaration-less XML plist as unknown and
broke the real-plutil `decodes xml plists that plutil cannot express as JSON`
CLI test. When `file` reports only text, read the source head once and label it
XML only when its first non-whitespace character (after a UTF-8 BOM) is `<`;
OpenStep and other text stay unknown.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(native): distinguish OpenStep data from XML source markup
Recognize XML markup instead of treating every leading angle bracket as XML.
Keep data-only OpenStep sources unknown, honor explicit JSON identification,
and decode BOM-marked UTF-16 prefixes before detecting declarationless XML.
Add source-format regressions and macOS CLI cases for data-only sources.
---------
Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Use one typed CDP resource-size owner in web and Electron capture. Supplied invalid sizes retain null plus completeness diagnostics; omitted sizes remain unknown without inventing malformed metadata. Preserve other resource facts and valid zero/positive values.
Validation: public CLI regression failed against baseline clamping; 36 focused controls passed after the fix, check:fast and docs:check passed. Fake CDP producer with real CLI/transport and tiny local files; no real Electron/browser or heavy profiling qualification.
Refs #4708
* fix(kuna): name the setting that owns a timed-out Kuna run
A Kuna run past its deadline surfaced the generic "Analysis took too
long. Try a smaller request" message. The request is already one
function or one whole-program scan, and the same run takes as long
again, so the advice was wrong and REA_KUNA_TIMEOUT_MS or
REA_KUNA_PROJECT_TIMEOUT_MS was never mentioned.
Each Kuna command now carries a KunaDeadline: the configured
milliseconds and the setting that owns them, from KunaConfiguration.
A timeout reports, for example, "kuna decompile did not finish within
REA_KUNA_TIMEOUT_MS (120000 ms)." Its remediation says to raise that
setting and restart REA, and for a single-function decompile that the
session's project export can still answer. AnalysisTimeoutError now
accepts a producer message and remediation, which projection uses and
per-waiter scan errors keep.
Refs #1957
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* docs(kuna): scope the provider_timeout statement to direct Kuna runs
A timed-out project export is reported through the export failure as
capability_unavailable, not provider_timeout.
Refs #1957
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(kuna): verify native xrefs under the project deadline
The provider's xrefs scan runs under REA_KUNA_PROJECT_TIMEOUT_MS, but the
native oracle in verify:kuna aborted at REA_KUNA_TIMEOUT_MS. A scan that
fit the project deadline but not the function deadline failed
verification although the provider answered it. Use the same deadline.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Read raw column metadata through the built-in PRAGMA in the private empty temp schema. Validate bytes before names become SQL identifiers, and select inventory subsets by exact encoded identity so unrelated malformed names cannot alias valid columns.
Validation: 5 focused identity controls; 15 selected producer/provider controls (33 skipped); public CLI baseline collision, fixed format refusal and valid-subset success; check:fast and docs:check. Generated inert SQLite fixtures on macOS arm64 with Node 24.18.0; no Windows qualification or large-file/stress workload.
Refs #4710
Reuse the shared validated XML document for depth and prototype-key preflight, preventing recovery-only values from being projected as observations. Preserve OpenStep and valid XML source semantics.
Validation: 46 focused checks; 5 final syntax controls; 10 encoding and public CLI/MCP overlap checks; check:fast. Public malformed-entity CLI changed from success to typed format failure. Generated inert artifacts only; no real Apple application or native engine execution.
Refs #4714
Validate XML bytes before shared syntax parsing so literal U+FFFD remains valid while malformed bytes and declaration mismatches fail at their producing boundary.
Validation: 92 focused checks, 3 selected fake Apktool boundary checks, 6 final encoding controls, check:fast and docs:check. Ghidra coverage uses tiny real files and descriptors, without a real engine or Windows qualification.
Refs #4712
Reuse the validated string inventory's numeric address order for O(log n + matches) exact reads, preserving every encoding at the selected address without a second lookup index. Keep streaming validation, session cache lifetime, full listings, provenance and caller choices.
Validation: complete latest-head CI passed. Original real Kuna 1.833 Linux correctness and warm-lookup measurements remain pinned to their recorded feature revisions in the PR. The latest local 300 ms export-deadline observation and unchanged 31-test provider recheck are explicitly documented there; no assertion or timeout was weakened.
Refs #4695.
Compose failed rollback with the original target or read failure, bind it to the exact workspace lease, and retain the failed attempt while cleanup remains unresolved. Remove the silent release wrapper and duplicate exception wrapping; keep clean failures retryable and unknown causes opaque.
Refs: #4704
Align malformed call-graph retry assertions with caller-attributed errors while retaining the parser operation in diagnostics and the cause. Keep all malformed shapes, corrected-response retry, exact native-command count and owned-client cleanup coverage.
Validation: clean main reproduces 3 failed / 14 passed; corrected complete Kuna boundary selection passes 266 tests across 26 files; full check:ci and complete current-head CI passed. Greptile 5/5; human owner reviewed the diff/evidence and explicitly declared readiness. Codex review remains unavailable due repository usage limits. Test-only; no production/schema/authority change or real-engine qualification.
Fixes#4701.
Reuse shared startup ownership for query waits and preserve opaque cancellation reasons, captured command facts and typed primary failures. Bind retained command errors to exact process and workspace leases; share close attempts and update earlier caller and aggregate errors after cleanup retry.
Refs: #4700, #4702
Reuse the shared startup owner for independent caller waits, final departure cleanup, and session closure. Carry the producer signal through version probing and fingerprint finalization. Retry cleanly cancelled acquisition after settlement while retaining authored failures and exact descriptor cleanup dependencies.
Refs: #4698
Restore test fixture runtime prerequisites and align assertions with current Evidence and uncertainty contracts. Separate actual CLI publisher refusal from explicitly admitted Hopper socket/CLI fixtures, retaining exact wire identity and cleanup coverage. Validate retained EVM Evidence against exact unchanged source-carrier bytes.
Validation: 88 targeted local test passes across three completed selections, two platform skips, full check:ci; hosted Windows probes and complete current-head CI. No production authority, authentication or schema change. Controlled fixtures do not qualify actual disassemblers or workers.
Fixes#4690.
Normalize shared installation and inventory failures at the session boundary, retaining producer operation metadata. Preserve exact cancellation reasons before admission and while waiting without adopting foreign cleanup authority.
Refs: #4689
Project the immutable shared installation failure through the existing authored-failure normalizer for each requesting operation. Preserve direct health identity, diagnostics, retryability and live cleanup dependencies.
Validation: ten baseline/candidate public CLI cases each; nineteen focused boundary checks plus four overlapping final cleanup controls, check:fast and docs:check passed. Concurrent callers share one acquisition failure without mutating its producer. No real Kuna, Windows, MCP-specific failure replay or performance measurement. Refs #4631.
Retain every focus declaration with its display header and identity, including explicit absence, unknown identities and duplicates. Derive singular summaries only for one unambiguous context; preserve scope and ambiguity through named CLI/MCP contracts.
Validation: original public CLI replay before/after, four failing baseline parser controls, 23 focused checks plus 11 overlapping provider/parser controls, converted MCP schemas, check:fast and docs:check. Bounded completed synthetic CLI warm reply tokens 282/513, elapsed 0.391/0.403s and CPU 0.622/0.622s; no efficiency claim. No live Android, vendor or large-output qualification. Refs #4660.
Send shell pm uninstall --user 0 through the shared quoted command boundary. Align tool guidance and Evidence limitations with the selected request scope while preserving device refusal diagnostics.
Validation: two provider controls failed before the fix and pass afterward; three public CLI/MCP scenarios capture exact arguments and equivalent success facts; nineteen command/SDK contract checks, check:fast and docs:check pass. No real device, multi-user data preservation, vendor/APEX behavior or performance measurement. Refs #4673.
Validate token items and their parameters before projecting exact declared policy tokens. Keep malformed and non-token values unknown; retain separate grammar for other policy headers.
Validation: real pinned Chrome public CLI baseline/candidate, 101 focused checks and 90 overlapping final checks, 834 upstream item-vector adaptations with two permitted refusals, check:fast and docs:check. Other browser/platform and policy-specific MCP delivery unverified; no performance measurement. Refs #4677.
Parse CDP repeated field lines independently before applying URI and quoted-parameter delimiters. Preserve declaration order and derive agent hints from retained links.
Validation: real pinned Chrome public CLI baseline and candidate each captured three responses per fixture; repeated Document links/hints restored and comma/single/Fetch controls unchanged. Three source controls failed before the fix. Sixty-one focused checks plus fifteen overlapping final network checks, check:fast and docs:check passed. No performance measurement or other browser/platform claim. Refs #4679.
Share validated interface Evidence assembly across success, cleanup failure and late cancellation. Preserve exact admission and completion abort reasons.
Validation: 26 service, MCP and host-refusal checks passed; check:fast passed. Thirty Linux x64 provider checks skipped on Darwin arm64, including expanded root-removal controls. Linux real-worker and public SIGINT reproduction remain unverified; no performance measurement. Refs #4683.
Attribute regex inventory failures to the accepted search operation while retaining wire identity, typed outcomes, captured facts and live cleanup dependencies. Preserve exact cancellation reasons at inventory and progress completion.
Validated with focused source, socket-boundary and public CLI fixtures, existing MCP contract checks, check:fast and docs:check. Simulated Hopper only; real-engine behavior and performance costs remain unmeasured. Refs #4680.
Carry producer cancellation separately from physical launch errors and diagnostic causes. Retain exact opaque reasons, LaunchSupport warnings, and rollback owners; consume the authored outcome without reclassifying failed launches from a later deadline.
Validation: focused client/provider/rollback and helper/MCP tests, check:fast, docs:check. Controlled native process ownership and Node bridge only; real Ghidra/JVM and Windows/Linux fault behavior unverified.
Refs #4517
Read bounded stable prefixes through the installation file owner, preserve unknown header diagnostics, and propagate cancellation or incomplete cleanup. Share extent and identity validation with full artifact reads.
Validated with 269 focused tests, check:fast, and docs:check. Controlled Darwin file faults reproduce the original lost owner; real Ghidra and native Windows/Linux verification remain unperformed.
Refs #4676
Reject special files without blocking, cap the properties payload at 1 MiB before allocation, and verify descriptor extent and the selected name. Preserve Latin-1 data, selected symlinks, physical filesystem facts, completed observations and exact failed-close ownership.
Reuse the stable bounded reader for profile artifacts and consolidate Ghidra file failure projection. Distinguish missing-at-open from disappearance or read failure, and attribute typed prerequisite failures at the consuming operation.
Validation: 384 focused tests across 14 files, 42 seed/extension controls, 32 final metadata/profile/CLI controls, check:fast and docs:check. Controlled macOS file/process fixtures; no real Ghidra/JVM or Windows/Linux lifecycle verification.
Bounded FIFO reproduction: baseline direct and public doctor required external cleanup around 2.2s; candidate returned independently around 0.4/0.57s, all fixture groups verified empty. Warm normal doctor baseline/candidate: elapsed 1.211/1.230s, CPU 1.239/1.247s, sampled family RSS 524/524MB, reply tokens 3116/3116. No performance improvement claimed; true cold and whole-agent costs unmeasured.
Refs #4672
Apply the existing cancellation projection to returned Results as well as thrown failures. Preserve the exact owner-close reason and keep shutdown joining the active resolution before later admission.
Strengthen the held-path regression to require typed cancellation instead of success. The original assertion and the stronger attribution control failed on unchanged cc2e18a26; focused validation and check:fast pass after repair. Controlled private files and recording provider ports; no native Ghidra claim.
Refs #4674, #4561
Replace synchronous Java execution with bounded owned commands. Keep task-scoped readiness observations fresh across later tasks, defer target client prerequisites behind a closeable owner, and join doctor cancellation and cleanup.
Preserve completed physical failures across late cancellation, exact cleanup dependencies and partial observations. Admit target/profile inputs before probing and return typed factory failures instead of unavailable client wrappers.
Validation: 405 focused tests across 24 files, 2 explicit skips; 109 provider/doctor controls rerun after final admission refinement; check:fast and docs:check. Controlled macOS Java-shaped commands and public CLI/MCP workflows; no new JVM/Ghidra or Windows/Linux lifecycle verification.
Measured warm public doctor baseline/candidate: elapsed 0.447/1.223s, CPU 0.624/1.237s, sampled family RSS 379/529MB, reply tokens 3133/3133. Ownership adds cost; no performance improvement claimed. True cold, whole-agent and MCP costs unmeasured.
Refs #4561. Remaining installation file admission tracked in #4672.
Retain physical stream errors and late caller cancellation independently. Expose an exact failed Darwin preparation root for cleanup retry without closing the shared reader.
Validation: 24 owned-command tests, 6 cancellation controls, 2 targeted Darwin cleanup retry controls, and check:fast. Controlled macOS processes only; no new real JVM or Ghidra verification.
Refs #4561, #4638
Carry asynchronous readiness through provider selection, session status, CLI and MCP. Join observations during cancellation and shutdown, keep status readable during finalization, and retain mixed provider errors and exact cleanup dependencies through the shared inspection failure owner.
Validation: 255 focused tests across 18 files; 21 registry tests rerun with primary restoration assertion; check:fast; docs:check and 5 README conformance tests. Registered MCP and SDK schema checks use controlled providers; no real Ghidra/JVM verification.
This is a prerequisite for #4561. The synchronous Ghidra Java producer and its doctor/client ownership migration remain unfinished.
BREAKING CHANGE: binary_session candidate availability can be unknown during owner finalization. Consumers must handle this state; the internal session status API is now asynchronous.
Refs: #4561
Preserve the owner-observed opaque caller reason through preparation, collection and cleanup failure wrapping without assigning cleanup authority or serializing the raw reason.
Validation: 68 focused process/deadline tests; check:fast; docs:check (5 conformance tests). Controlled local process fixtures only; no native provider verification.
Refs: #4638
Produce typed cancellation at each command checkpoint while retaining the exact opaque reason as cause. Preserve independent typed physical failures without downstream equality guesses.
Validated 72 focused tests across nine files, static checks, docs and five conformance controls, including real client/service admission races and matching physical-error controls. No native Cutter engine or new public-transport reproduction claimed.
Refs: #4635
Parse command status and validate session identity at one bridge boundary. Keep the preflight watermark requirement separate from exact completion receipt proof, and retain status as a compact auxiliary diagnostic.
Reuse authenticated protocol and descriptor fixtures. Validate 66 focused tests across eight files, static checks, docs and five conformance controls. CLI refusal controls use an authenticated loopback fixture; native Cutter was not verified.
Refs: #4607
Attribute shared startup failures separately for each accepted operation. Preserve the original stage as a compact diagnostic, with typed timeout budgets, limits, output, recovery and exact cleanup dependencies.
Link cancellation wrappers to the originating live cleanup owner, including release before and after wrapping. Remove the redundant typed-error branch.
Validated 130 focused tests, static checks, docs and five conformance controls; four source controls rerun after fixture type correction. Public CLI uses a loopback supervisor; no native IDA engine verified.
Refs: #4614, #4633
Retain the primary error through cleanup composition and resolve its presentation only after live ownership confirms release. Preserve historical diagnostics separately from producer facts.
Validate domain primary types, nested owners, denied retry and late observations; exercise public CLI and MCP projection plus firmware descriptor retry. Focused tests, source controls, static checks and docs pass. Native providers were not verified.
Refs: #4628
Classify the WABT reader file-size boundary as a deterministic resource constraint with its input allowance. Give Go input-size refusals selected-input recovery rather than evidence-write advice. Preserve configuration and cleanup semantics.
Validation: 66 focused tests across four files including public CLI refusals, plus docs checks and five conformance tests. Sparse admission fixtures and version-only WABT stubs; no real WABT execution. Refs #4621, #4622
Parse window focus into declaration records, normalize explicit null references once, and preserve original named lines in Evidence. Keep fallback based on declaration presence and retain unknown focus when both sections omit fields.
Validation: 86 focused tests including public CLI/MCP with advertised schemas, plus docs checks and five conformance tests. Fake ADB only; no physical device verified. Refs #4608
Share source identity and one iterative pattern walker across declarations and assignments. Traverse nested default targets without losing dependency paths, preserving existing scalar-default occurrence identities and conservative unresolved slots.
Validation: 405 focused tests across five files, including public CLI dependency Evidence. Refs #4609, #4604
Use the canonical cooked literal value for for-in and for-of string cardinality, preserving possible assignments and prototype uncertainty across quoted and template spellings.
Validation: 473 focused tests across six files, including public CLI assignment Evidence and exported return shapes. Refs #4617
Do not interpret exact caller cancellation reasons as descriptor ownership records. Preserve checkpoint-shaped reasons through successful close retry and supply cancellation context in direct signal-aware cleanup consumers.
Validation: 156 focused tests across 14 files passed; no native provider execution. Refs #4610
Project canonical timeout and cancellation facts at the provider boundary while preserving causes, partial observations, live cleanup dependencies and provider-authored deadline guidance.
Validation: four real SDK/HTTP catalog and tool timeout controls fail before the fix. After rebasing onto the Ghidra timeout guidance change, 189 tests across ten files pass, including public CLI/MCP workflows and affected Ghidra tests. Typecheck/lint, docs, format and discovery pass. No native JEB engine or target verification. Refs #4597.
Bind pending SDK resources to exact cleanup leases, coalesce disposal, and retry only unresolved children. Shared startup callers retain final cleanup while independent peers can cancel their waits.
Attribute SDK aborts by exact owned rejection identity, preserve physical failures under late cancellation, and retain completed Evidence at the application boundary.
Validation: 83 focused tests across eight files; typecheck/lint, docs check, formatting and discovery. Controlled real HTTP/SDK faults only; no native JEB engine or target verification. Refs #4603.
