Prove total orders with Kahn traversal and use ranks; stream forward and reverse reachability for general partial orders instead of retaining every closure. Aggregate unordered conflicts by declaration while preserving missing-pair counts and independent malformed-input diagnostics.
Validation: 33 focused trace domain and CLI/MCP boundary tests passed. Six bounded Flameox executions preserved answer digests and cycle rejection; combined median peak RSS fell from 391.13 to 127.72 MiB. General partial-order traversal costs more CPU; the experiment records that tradeoff. Full suite and real-provider lanes were not run.
Require terminal stderr before launching the automatic setup wizard, matching the CLI's interactive admission rule. Print the manual setup command when diagnostics are redirected.
Exercise a real PTY with stderr redirected to a fixture file and verify the installed command is not launched for setup. Installer lane: 33 tests passed; targeted regression, typecheck/lint and npm pack dry-run passed. Full package verifier was not run under the requested focused-check resource policy.
Use iterative cycle detection, memoized reachability and compact unordered-group membership. Preserve declaration diagnostics on cyclic input while skipping checks requiring a valid partial order. Report duplicate group declarations once instead of repeating pair errors.
Replace redundant source cycle/missing-pair assertions with public CLI/MCP coverage including 6000 events, dense not-before constraints and a large unordered group. Validate acyclic branching and mixed relation semantics. Document bounded Flameox CPU/latency gains and higher observed peak memory; agent/model costs remain unmeasured.
Close persisted Evidence, bundle, profile and unknown record shapes so parsing cannot silently remove caller metadata before checking semantic identity. Keep constructor provider projection explicit and JSON payload fields open.
Exercise actual MCP export/import with independent record and manifest mutations, unchanged state after rejection, and a valid round trip. Prune duplicate source assertions. Validation: focused domain/CLI/MCP/filesystem checks, converted JSON Schemas, docs and typecheck/lint.
Bind the owned demo helper to REA's private stdin stream and unwind child
cleanup on EOF during startup or analysis. Verify graceful and SIGKILL owner
exit through real CLI/MCP workflows and preserve unrelated processes.
Assisted-by: Reimu (gensokyo-skills:reimu-incident-triage)
Summarize saved Wireshark4.x captures with enforced native resource limits, serialized admission, stable snapshot verification and retained fact/text pages. Preserve complete inventories within resource budgets and source uncertainty; report cleanup failure with partial Evidence and retained ownership.
Verified six recorded corpora /13026 packets with independent Scapy/Python oracles, real CLI/MCP/schema parity,60 SNI names,132 DNS addresses and1000 URIs. Real TShark CI now exercises summary/paging. Actual Windows/Bun Wireshark remains unverified; macOS capture requires a separate enforceable-bound decision.
Co-authored-by: 3QL <85309073+3QL@users.noreply.github.com>
Co-authored-by: N0zoM1z0 <161784452+N0zoM1z0@users.noreply.github.com>
Follow-up to #2345; thanks to @gefrezim for the import-time home lookup diagnosis.
Normalize only an empty Windows USERPROFILE before runtime dependency imports, preserving all nonempty caller selections. Reuse the existing public setup dry-run regression in the Windows curated CI lane. No additional filesystem query, engine call, cache or retry.
Verified real Windows Node 24.18.0 CLI unset/empty/selected-profile controls, setup dry-run home selection and unchanged configuration, and baseline/candidate SDK stdio MCP catalogue parity. Nine related tests and static/docs checks passed. No IDA or Ghidra engine was used. Scoped resource observations and measurement limits are recorded in experiments/analysis-efficiency/windows-empty-profile.md.
Fixed the review finding and original CI package-scope ownership failure. Current-head CI is green; the unrelated MCP identity timeout passed on a failed-job rerun without changing its code or timeout. Current-head review completed with no further findings.
inspect_macho launched otool, nm, two dyld_info reads, dwarfdump and
vtool one at a time. Each launch also resolves the tool through
`xcrun --find`, so a warm call spent about 350 ms in eight serial
supervised launches. The slice commands only read the selected
artifact, so launch them together. Every launch still settles before
the result, and declared order still fixes provenance and which
failure is reported.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
When a pwntools or pwndbg stage cannot confirm that its owned process
group stopped, the cleanup_incomplete result listed only the run ID and
runtime directory, although a descendant can still be live. Add the
retained supervisor's pid and process-group coordinates, as
OwnedCommand reports for a non-transferred cleanup failure.
Fixes#2274
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Reuse the domain ZIP format list for session schemas, verify full Harmony
CLI/MCP lifecycles, and correct native view guidance and audited docs drift.
Refresh locked dependency notices and gate inventory/provenance consistency,
including notice inputs in the cached documentation gate.
Fixes#2454
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Reuse the current advertised binary_session definition for real SDK reply
validation while preserving all inventory assertions and the 10-second deadline.
Avoid initializing validators for unrelated tools in this correctness check.
Closes#2453.
Assisted-by: Reimu (gensokyo-skills:reimu-incident-triage)
Fixes#2257. Thanks to @chqcjq for the native self-test retry diagnosis in #2266.
Preserve retry disposition through analysis workflows. Native warm admission requires exact sharing/lock codes and affirmative pre-allocation evidence; cleanup ownership takes precedence. No automatic retry or new analysis cache.
Verified with real Windows 11 CLI, same-connection SDK MCP normal/conflict/recovery, the Windows native verifier, 99 focused tests and static/docs gates. Official npm 11.16.0 launcher sources were the application input. No IDA or Ghidra engine was used for this change. Win32 5 was not reproduced; no resolution claim for #1701/#2196. Performance measurements and limits are recorded in experiments/analysis-efficiency/windows-runtime-retry.md.
Add explicit MCP Dart AOT summary delivery and exact retained string-pool pages while preserving complete defaults, CLI behavior, ABI selection, original Evidence identity, partial coverage and lower bounds.
Real LocalSend SDK replies shrink 92.05% with approximately unchanged SDK time. Two actual Codex pairs preserve all selected facts and reduce total input 48.75% / 42.38%, uncached input 33.63% / 42.52%, and task time 23.95% / 30.30%; all client errors/repeated calls remain counted. Preserve old portable Evidence and the real CLI complete result after rebasing onto SAZ/qlog main. Curated results keep measured and integration runtime pins separate; raw artifacts stay outside Git.
Refs #2016.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Inspect bounded recorded raw HTTP/XML SAZ archives and page retained capture Evidence. Preserve exact stored bytes and provenance, validate ZIP/XML integrity, and require an enforceable decoder memory bound.
Verified 356 real recorded sessions against independent byte and metadata oracles, public CLI/MCP parity, retained pages, and related interface checks. Current Fiddler Everywhere GUI and non-Linux producer workflows remain unverified.
Co-authored-by: zahhak <1792039+zahhak@users.noreply.github.com>
Co-authored-by: N0zoM1z0 <161784452+N0zoM1z0@users.noreply.github.com>
Add explicit package summary delivery and exact retained manifest pages while preserving complete MCP defaults, one-shot CLI output, provider requests, source identity and session lifetime. Return distinct linked Evidence and preserve partial/null facts; reject unavailable retention before engine work.
Validation: 27 relevant SDK/domain/Android boundary checks, focused test-only correction rerun, full supported compiler/lint and docs checks, real NewPipe SDK pairs, authentic old Evidence continuation, four actual Codex trials with failures/regressions retained, and all 39 exact-head CI checks. Package replies shrink 119,527 to 8,659 bytes; no model-token or time gain is established.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Add opt-in, owned payload-completion cgroup witnesses without changing default receipts, terminal-counter semantics, resource limits, or REA public APIs. Preserve unknown and nonzero outcomes; verify ownership, observer capacity, cleanup, and actual SDK/model workflows.
Validation: eleven Python semantic tests, supported full-project typecheck and lint, controlled Linux failure/cancellation/capacity probes, real managed/WASM/Android SDK and Codex runs, and all 39 exact-head CI checks. Curated evidence retains mixed timings and whole-scope uncertainty.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Scope source-file cleanup fault injection without poisoning enclosing Git
metadata reads. Cover both environments and verify that the retained native
handle stays live through retry, then closes successfully.
Closes#2428.
Assisted-by: Kogasa (gensokyo-skills:kogasa-surprise-testing)
Verify the static EVM descriptor operation, CLI command, factory and input/output schema bindings with compiler-backed negative assignability checks. Prove SDK duplicate registration and invalid target input fail before provider invocation, preserving named bindings and separate lazy owners.
Validated complete supported single-thread typecheck, focused12 SDK/composition cases plus a targeted2-case provider-spy rerun, static checks, checked pre-push, manual review and all exact-head CI checks. Production behavior/defaults and provider identity remain unchanged. Dynamic third-party module admission and ABI negotiation remain deferred.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Add optional owned Codex tool/usage capture and exact-name discovery experiments while preserving default ephemeral trials and all REA public APIs. Retain scoped managed and actual Android results, mixed outcomes, source identities and resource unknowns.
Six semantic Python tests, full checked pre-push, byte-identical own-file rebase verification, manual review and all 39 exact-head CI checks pass. Historical main fixture failures and their independent repair remain documented; measured source/runtime pins are unchanged.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Reject native callee successes without an explicit truncation marker while preserving provider failure reasons. Verify positive procedure inventory/search and function-result parity in the real-provider lane; support independent string/xref oracles and label their absence as unverified.
Verified with Windows IDA Pro 9.3: eight normal production CLI/MCP checks passed, and six real-provider response-fault cases reproduced old false positives and validated the correction. Windows IDA Free was not used. No personal paths or license registration information are included.
Repair inconsistent complete-file Ghidra test probe sizes after the shared ELF-header validation change, deriving sizes from the fixture buffers. Reproduced the three failures before the correction; 115 related installation/doctor and ELF tests pass afterward. Runtime parsing and installation behavior are unchanged; no real-Ghidra verification is claimed.
207 focused IDA tests and static/documentation checks passed. Automated review completed on d1e0b1095 without findings. CI run 38327862093 passed all 36 jobs, including Windows, all six test shards, coverage/timing, and real Apple CLI/MCP workflows.
Follow-up to #1544.
@modelcontextprotocol/client 2.3.1 declares callTool(params, options).
Three verifiers still used the v1 shape callTool(params, undefined,
{ timeout }), so the third argument was dropped and those calls ran
under the SDK's 60 s default instead of the 150-180 s they request.
Pass the options second.
Use the shared resource scope and private runtime root for signature snapshots. Retain failed file-handle/root cleanup, preserve primary failures alongside cleanup diagnostics, and wait for binding operations during client close before retrying owned resources.
Validate root cleanup failure, blocked retry, repeated client close and in-flight acquisition. Existing provider/MCP/CLI tests and the real Darwin codesign ancestor-swap case pass. Dedicated native file-handle close-failure injection remains unverified.
Keep complete default delivery and exact filters. Select stable ID pages at the ledger owner before cloning, return filtered totals and continuation offsets, and preserve detached records through the shared session port.
Validate public MCP paging, safe-integer edges, snapshot replay and converted JSON Schemas. Document the bounded Flameox workflow comparison and its token/CPU/latency measurement limits; no real model performance claim.
Validate ELF identification/header versions and declared header bounds at the shared owner. Reuse parsed facts in target selection, Go, Dart and MIPS readers while preserving their stricter format policies and x32 support.
Repair synthetic ELF fixtures and remove a superseded Go error-text assertion. Validation: 74 source tests, 202 filesystem/Go/MIPS boundary tests, typecheck and lint. No new real-provider claims.
Add caller-selected managed metadata summaries and original collection pages while preserving complete CLI/MCP defaults and previous-runtime Evidence. Real SDK replies shrink substantially; actual Codex token/latency results remain mixed and are documented without a token-saving claim.
Validated focused SDK/domain tests, full static/docs checks, real two-assembly repeated workflows and portable old Evidence, real SQLite WASM, and all exact-head CI checks. Raw experimental artifacts remain private.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
* fix(javascript): resolve module-relative Worker URL references
* fix(javascript): validate Worker URL paths independently of suffixes
* fix(javascript): preserve directory-form Worker URL paths
* fix(javascript): unwrap type assertions in Worker URL references
* fix(javascript): validate Worker bindings before URL resolution
* fix(javascript): leave drive-letter Worker URLs unresolved
Use WHATWG file URL parsing before pathname decoding to detect drive-letter
forms whose resource identity depends on the unmodeled source URL scheme.
Preserve encoded and ordinary relative pipe paths, and cover false matches
through the compiled CLI and resolver boundary tests.
* fix(javascript): preserve Worker URL identity boundaries
Recognize known Worker bindings before filtering callee spellings so ESM
aliases and transparent constructor wrappers reach URL analysis. Keep
root-relative module URLs unknown without a modeled runtime base root.
Index direct source-local global writes with the shared mutation collector
before promoting URL and browser Worker constructors. Preserve lexical
shadowing and unrelated property identities, and exercise all three review
regressions through the compiled CLI.
* fix(javascript): invalidate mutated import.meta URL bases
Track source-local import.meta writes alongside ambient global writes so
Worker URL expressions keep an unknown base after a possible URL mutation.
Tag the identity roots independently and cover direct mutation and unrelated
property cases through the compiled CLI.
* fix(javascript): close Worker URL identity and pathname gaps
Resolve ambient object identities before appending written properties, and
retain possible source-local aliases across rebinding. Reuse intrinsic
invocation handling for aliased and bound mutators, array spreads, and
transparent TypeScript wrappers while preserving Electron write semantics.
Keep malformed-percent nested Worker URL pathnames unknown because the base
scheme is unavailable and Node rejects these file URLs. Preserve tolerant
HTML URL handling, valid escapes, and query/fragment evidence.
Cover the new cases and unrelated writes through the compiled CLI with
existing-file decoys. All 591 targeted tests and npm run check pass.
* fix(javascript): bound alias expansion and classify Worker paths
Join ambient identity projections instead of enumerating alias histories.
Memoize repeated value and provenance queries while preserving capture,
depth, and cycle-dependent uncertainty.
Keep Node Worker string arguments unresolved without filesystem or cwd
context, and admit qualified global URL constructors through identity checks.
Validate three 40-level alias branch shapes through the compiled CLI,
925 source tests, 25 CLI/boundary tests, and the full npm run check gate.
* fix(server): preserve partial Evidence target-content provenance
Pass the exact named contract returnsTargetContent flag through the shared partial-Evidence error delivery path introduced by the shortlist integration.
Exercise successful and failed recording through in-memory MCP for marked and unmarked contracts, preserving source payload and original failures. Observe both shared-handler emission and wire delivery so the SDK error-marker fallback cannot hide omissions. Suppressing the flag fails exactly the two marked cases.
Validation: check:fast, check, 736 semantic tests, 29 MCP transport/error tests and six compiled static Worker CLI journeys pass. No live target/provider execution or hosted publication.
* fix(javascript): structurally share semantic dependency proofs
---------
Co-authored-by: baron <richstyles@gmail.com>
* fix(native): honor Mach-O chained pointer locations and library ordinals
Walk declared page starts and next links before decoding chained pointers,
preserving plain pointers between fixups, zero-offset rebases, and known
facts in malformed chains. Interpret positive dylib ordinals with dyld's
special-ordinal cutoff so valid dependency provenance is retained.
Give the Swift verification fixture an explicit module name so link-mode
filenames do not produce invalid identifiers.
Validation: 46 focused regressions; pre-fix reproductions in an isolated
copy; real macOS CLI/MCP verification of 10 compiled Objective-C/Swift,
arm64/arm64e, chained/legacy and symbols/stripped fixture variants.
* perf(native): reuse parsed Mach-O chain membership
* test(native): validate parsed-chain repair against local gates
* test(native): pin chained fixup page failure provenance
* fix(native): bound parsed chained page facts and metadata
* test(server): pin shared partial Evidence provenance
---------
Co-authored-by: baron <richstyles@gmail.com>
Verify unchanged real EVMole worker replies through compiled CLI and stdio MCP, including late SDK cancellation, permission-induced retained cleanup failure/retry, preserved malformed-input diagnostics and a healthy real SQLite WASM peer. Add launcher receipts to prove blocked retries do not start another EVM worker.
Preserve production contracts/defaults and explicit frozen-runtime/platform/measurement boundaries. Final head cf558a0661eb29bda932242abc3687b1860428e7 passes all 36 CI checks (run38313631216); manual final review and related real Factory rerun pass. Raw artifacts and receipts remain outside Git.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
* fix(browser): compare PNG IHDR bytes without ASCII folding
Screenshot checks already compare the PNG signature as raw bytes, then decoded the chunk type with an encoding that maps high-bit bytes onto IHDR.
Co-authored-by: Cursor <cursoragent@cursor.com>
* fix(browser): satisfy the strict index check in the IHDR test
Co-authored-by: Cursor <cursoragent@cursor.com>
* refactor(browser): share byte-faithful PNG header facts
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: REA preparation lane <rea-preparation@localhost>
* fix(artifacts): bound archive decompression by source size
A ZIP/APK/IPA/MSIX member is attacker-controlled and can expand far
beyond the archive that carries it. Inventory hashed every member to EOF
and extraction wrote every member with no ceiling, so a ~2 MiB archive
could drive ~30 s of CPU during inventory and ~150 s plus 2 GiB of disk
writes during extraction.
A fixed per-entry byte cap is the wrong shape: it rejects legitimately
large members inside large packages. The ceiling instead scales with the
archive's own size (500x, with a 256 MiB floor), so large packages keep a
proportionally large budget while a small archive cannot amplify into
unbounded work. The cumulative declared size is charged from the central
directory before any entry is decompressed, and zip.js already verifies
streamed output against each declaration, so the check is exact and
covers both inventory and extraction through the shared reader.
* fix(artifacts): own decompression accounting per traversal
* test(artifacts): verify declared ZIP output enforcement
---------
Co-authored-by: baron <richstyles@gmail.com>
morluto
Ysopwnable
Quentin Limo
Kaoru0822
Luis
Ilian Iliev
Aurelio Bernal
Yann Roberto
Yijie Lu
Faheem Ahmad
Dery Ferdika
The one -Niu