Propagate CLI/MCP cancellation through shared Cutter discovery, stop queued probes, and drain admitted workers. Deliver terminal bridge results only after socket closure. Preserve confirmed command Evidence and caller reasons when cancellation arrives during finalization.
Validated 66 focused tests, full CI static/docs gates, actual CLI SIGINT, compiled stdio cancellation/recovery, and official Linux Cutter 2.5.0 Qt5 CLI/MCP ELF workflows. Native UI cancellation and descriptor filesystem cancellation remain unverified.
Fixes#3946.
* fix(application): retain completed evidence through cancellation
Validate completed provider reports before classifying cancellation and preserve Evidence and captured diagnostics. Share MCP recording and delivery across custom observation handlers. Add source-owned completion and public SDK recording controls. Refs #3923, #3927, #4261.
* fix(wasm): retain final observations through cancellation
Keep completed Evidence after finalization and through CLI/MCP delivery. Preserve output and cleanup failure precedence. Refs #4254.
* fix(browser): retain final observations and forward cancellation
Classify final CDP detach and completed scenario cleanup at the provider boundary. Translate completed partials into canonical Evidence for MCP recording, and forward CLI cancellation to passive Electron tools. Refs #4255, #4256, #4257.
* fix(ghidra): isolate cancellation of shared startup waiters
Keep startup group-owned while each caller has an independent abortable wait. Await final-owner cleanup and preserve cleanup facts before retrying. Reuse the shared abortable promise helper. Refs #3983.
* fix(mcp): keep shutdown registrations until cleanup succeeds
Retain repeated event requests while unresolved owners need cleanup, and unregister only after success. Verify coalescing and failed-owner retry. Refs #4253.
* fix(adb): own transfer snapshots and stage screen publication
Bind transfer digests to the exact private input bytes and preserve caller APK names. Retain failed descriptors and private-root owners with typed cleanup evidence; publish captures exclusively after the cancellation cutoff. Refs #4251, #4252.
* fix(verification): report ADB cleanup failures
Always close after attempted verification and keep unexpected primary failures and cleanup details visible. Exercise the production script with controlled imports, including null and undefined throws. Refs #4260.
* fix(runner): stop owned Windows commands on cancellation
Use the retained ChildProcess handle for IPC cancellation and parent disconnection. Keep the SQLite lock until observed child close, and include native Windows regressions in the curated CI lane. Native Windows results remain pending. Refs #4258.
* fix(website): canonicalize link containment boundaries
Compare resolved targets with resolved repository and site roots. Accept directory aliases while preserving escape checks and keep fixtures within their temporary owner. Refs #4264.
* docs(testing): cover completion races and cleanup retry
Clarify validation-before-cancellation, independent startup waiter controls, producer and adapter coverage, and repeated shutdown events.
* fix(filesystem): retain file owners and completed reads after close failure
* fix(session): report retained snapshot staging in cleanup status
Fixes#4277. Include both snapshot and profile file-resource scopes in the owning status fact, preserving publication receipts and cleanup retry.
* docs(mcp): clarify recovery of raw partial observations
Document the observed SDK cancellation boundary and the open design decision in #4276 without promoting raw errors to Evidence.
* fix(kuna): own the complete installation inspection lifetime
* test(kuna): cover session cleanup of shared cache resources
* fix(hopper): isolate publisher operation resource admission
* fix(hopper): preserve launcher cleanup failure classification
* fix(adb): retain ordered cleanup and screen publication outcomes
* fix(runner): retain Windows job ownership through cleanup retries
* docs: clarify inspection admission and ordered cleanup regressions
* fix(ghidra): preserve startup cleanup and request deadlines
* fix(runner): supervise stdin errors and bootstrap native controls
* docs: require queued lifecycle controls to survive admission
* fix(cli): leave cleanup resolution with resource owners
* test(adb): verify owned install snapshots through CLI
* fix(errors): exclude resolved primary cleanup resources
* fix(adb): resolve publication cleanup after its handle closes
* fix(android): resolve JADX hash cleanup after owned close
* refactor(android): share hash failure cause options
* fix(kuna): gate commands on retained descriptor cleanup
* docs: cover cleanup admission and owner-confirmed recovery
* refactor(artifacts): bind cleanup errors to confirmed release
* fix(adb): resolve recovered cleanup at retained owners
* fix(kuna): resolve installation cleanup after owned close
* fix(kuna): gate cache reads on retained hash cleanup
* fix(setup): preserve publisher cleanup admission failures
* fix(android): resolve the reported retirement cleanup error
* test(runner): scope post-spawn fault control to POSIX
* test(adb): cover persistent snapshot owner retirement
* fix(kuna): preserve file outcomes with explicit cleanup ownership
* fix(artifacts): resolve errors through retained cleanup owners
* docs(lifecycle): carry cleanup ownership in typed outcomes
* fix(adb): bind staged screen errors to their cleanup lease
* fix(kuna): bind profile cleanup through a close capability
* fix(artifacts): retain descriptor resources in cleanup errors
* fix(pcap): bind existing errors to snapshot retirement
* fix(windows): load source owner through its file URL
* test(build): copy the command runner helper into fixtures
* fix(mcp): own shared observation providers in the transport
* fix(jadx): resolve every retained cleanup retry error
* fix(kuna): bind retry errors to every retained cleanup owner
* fix(adb): bind admission and shutdown errors to exact cleanup leases
* fix(windows): resolve command paths before native ownership
* test(publisher): assert exact descriptor recovery without release metadata
* docs(lifecycle): require every retry error to follow its owners
* fix(kuna): keep hash cleanup state stable across concurrent retries
* test(windows): exercise surviving descendants with scoped shutdown deadlines
* fix(uninstall): preserve primary read errors through cleanup failure
* test(publisher): assert cleanup Result and exact owner recovery
* test(windows): observe supervisor resources through IPC disconnect
* fix(browser): retain typed cancellation after interrupted actions
* test(publisher): assert ownership phases without retry-count coupling
* fix(lifecycle): preserve abort reasons through shared provider execution
* fix(jadx): preserve execution and late hash cancellation causes
* fix(adb): compose late cancellation with exact cleanup failures
* test(mcp): use the shared budget for session workflows
Own CDP transports from admission through observed close, retain failed
owners for retry, and keep active peers separate from pending cleanup.
Preserve primary errors and completed facts through final cancellation.
Forward passive Electron CLI cancellation and dispose browser, runtime and
scenario owners from CLI and MCP shutdown.
Refs #4255, #4256, #4257
Validate completed reports before cancellation projection, retain valid
Evidence and captured command diagnostics, and convert raw cleanup partials
at their application owner. Record partial Evidence consistently across MCP
handlers. Keep page inspection separate from unfinished bundle analysis.
Refs #4261
Admit all four derived investigation workflows until Evidence recording settles, so close drains earlier work before clearing the ledger. Honor cancellation after awaited terminal progress before publishing a result.
Refs #4002, #4174
Co-authored-by: Swatantra Yadav <maverickswatantra@gmail.com>
open_binary read snapshot_path before it entered the session's lifecycle
order, so a current_document or close_binary dispatched after it could
run first: the read returned Evidence for the previous target and the
close left the snapshot target open. The session now loads the snapshot
inside the open's own turn.
Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com>
Use shared recorded-partial delivery on start, load and instrument errors. Return a retained reference only after a successful write; preserve the original inline partial when retention fails or is unavailable.
Include the selected output advertisement profile while retaining canonical output validation in the identity. Reuse the SDK output projection and select doctor expectations from both profiles. Correct the compact input error documentation.
Measure actual wire result JSON, including its provenance marker, without allocating the full serialization. Retain complete SDK diagnostics through the existing Evidence writer and return a bounded reference or explicit retention failure. Keep fitting escaped errors and subsequent valid calls usable.
Validation: actual SDK client/transport retention readback and fit/one-byte-over delivery boundaries.
Fixes#3221
Related: #3267
Co-authored-by: CRYPTONIKAV <29835619+CRYPTONIKAV@users.noreply.github.com>
Probe the selected clients through their owning providers and expose current host availability through binary_session. Keep device/serial and target applicability separate. Reject unrecognized successful version output; retain the bounded producer output and original version spelling. Apply the documented Apktool floor, including the verified Debian dirty build.
Validation: provider/version tests and public MCP grouped availability; no real Android device or APK provider verification claimed.
Fixes#3067Fixes#3068
Related: #3077
Co-authored-by: Taksh <takshkothari09@gmail.com>
Capture the current entrypoint once without inventing a path. Preserve unknown registration ownership through setup and doctor, refuse unsafe registration plans, and retain exact pinned-launcher readiness checks. Hash the same full or compact schema projection advertised by each MCP server and use that identity in doctor comparisons.
BREAKING CHANGE: server.command_path may be null when the invocation path is unavailable. Registration and setup statuses, and doctor classifications, may report unknown.
Reviewed-PR: #3230, #3015
Co-authored-by: qorvr <330186599+qorvr@users.noreply.github.com>
Co-authored-by: Hemang Varshney <218273197+hemv-857@users.noreply.github.com>
Move the shared input-error constructor and operation identity to a leaf contract. View leaves no longer import their runtime composition owner. Replace a duplicate graph assertion with compiled CLI summary/facet coverage for omitted roots and unresolved hierarchy endpoints.
* perf(managed): add retained member summary and pages
inspect_managed_members always returned the assembly-wide member_refs, so
one selected or unmatched type produced megabytes, and an unselected
ordinary assembly exceeded the MCP response budget. Add opt-in
detail: "summary" that retains the complete Evidence and returns identity,
selection with unmatched names, coverage, user-string coverage, limitations
and per-collection counts. Page original rows through inspect_analysis_view
with kind managed-members; each page names its collection so no row schema
is chosen by union order. Complete delivery stays the default.
Refs #3010
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* perf(managed): validate a sealed member parent once across pages
Each managed-members page reparsed the whole parent report, about 35 ms
for an 18 MB assembly result regardless of the page size. The analysis
view service now reports whether the parent Evidence is a sealed
snapshot, and the member view reuses the validated report for that
sealed result through a weak reference. Unsealed input is still
validated on every call.
The round-trip test fixture now carries an exported type, an ldstr
literal, a field access and a call edge, and requires every collection
to be nonempty so a row schema that drops facts cannot pass on empty
arrays.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(managed): keep delivery detail out of retained Evidence identity
Summary, omitted and complete delivery of the same managed inspection
now retain one parent record, as inspect_macho already does.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Use one named input-schema projection for clients and catalog identity, preserve dialect declarations, and bound compact schema depth. Keep canonical SDK output validation when output advertisements are disabled. Release request-owned state on cancellation and carry upstream limitations through composed workflows. Preserve doctor full-output envelopes and validate navigation facts at their owner.
Reviewed fixes: #2088, #2108, #2513, #2684, #2693, #2695, #2836, #2864, #2868, #2877, #2933, #2998, #3002, #3015.
Validation: focused schema, SDK output-validation, context, and workflow tests passed; no schema dialect or canonical output validation was relaxed.
* fix(frida): drain admitted operations and honor queued cancellation
* fix(frida): settle cancelled loads before queued resume
* test(frida): share lifecycle fixtures without duplicating assertions
* fix(frida): preserve operation admission before session close
* test(apple): bound cold archive fixture compilation
* fix(frida): retain cancelled script creation evidence
* test(release): scope preflight schema to selected jobs
* fix(process): report no remaining time after a provider deadline fires
Node timers can fire while performance.now() is still a fraction of a
millisecond short of the monotonic deadline. ProviderStartupDeadline then
reported interruption "timeout" and an aborted signal while remainingMs()
returned ceil(0.x) = 1, so polling loops guarded by remainingMs() > 0
(Hopper and Ghidra startup) could run past an elapsed deadline, and the
real-timer test "aborts a bounded wait when its real monotonic deadline
expires" failed intermittently in CI (expected 1 to be 0).
Once the deadline's own timer has classified a timeout, report zero
remaining time. A deterministic regression test fires the timer with the
monotonic clock 0.4 ms short of the deadline.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(comparison): recheck historical calls-only changes (#2223)
* fix(comparison): preserve incomplete function call evidence
Keep missing, unknown, partial and unresolved call relationships unknown in function comparisons and reconstruction claims. Aggregate IDA caller and callee coverage without overwriting uncertainty.
Validate with 144 focused tests, typecheck/lint, docs checks and test discovery. Real providers were not verified; the full headless IDA session suite requires unavailable Windows native controls.
* fix(comparison): recheck historical calls-only changes
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
(cherry picked from commit 688defca0f)
* fix(ci): isolate Linux Hopper process fixture inventory
(cherry picked from commit 8cac51e121)
* fix(android): make DEX descriptor pattern portable
(cherry picked from commit 1335bcddc5)
* fix: repair the static-check failures on the default branch (#2895)
* fix: repair the static-check failures on the default branch
The default branch fails its own required `Static checks` job three ways, so
every pull request reports that job as failed regardless of its own change.
- `npm run format:check` reports `src/domain/functionComparison.test.ts` and
`src/ida/IdaOperationRunner.ts`.
- `npm run lint` reports a duplicate `import { join } from "node:path"` in
`tests/boundary/mcp/ghidraInventoryMcp.test.ts` (lines 4 and 17).
- `npm run typecheck` reports `format: "devin"` missing from the client format
union in `src/application/SupportedClients.ts`, and an unchecked index in
`tests/boundary/cli/clientProfileLifecycle.test.ts`, whose pair array is
inferred as `string[][]`.
Apply the pinned formatter, drop the duplicate import, list the `"devin"` format
that its own definition already uses, and mark the pair array `as const`.
Validation: `npm run check:fast` and `npm run format:check` pass.
* fix(setup): keep Devin on the JSON configuration dialect
---------
Co-authored-by: 吴佳浩 <wujiahao@supremind.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
(cherry picked from commit df3d96da80)
---------
Co-authored-by: morluto <morluto@users.noreply.github.com>
Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Co-authored-by: Lucien <164857476+dakjdakd@users.noreply.github.com>
Co-authored-by: 吴佳浩Alben <35291372+XiaoBinGan@users.noreply.github.com>
Co-authored-by: 吴佳浩 <wujiahao@supremind.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
* feat(native): read embedded cargo-auditable Rust dependency lists
* fix(native): trim PE section padding and compute dependency cycles iteratively
* fix(native): ignore SHT_NOBITS sections and validate dependency edges linearly
* fix(native): bound compiler path scanning and retained hashes
* fix(rust): pass package arrays as named parameterized test cases
* fix(rust): admit graph expansion and preserve resource units
* fix(rust): reject malformed packages, false cycles, and hidden compiler paths
Fail the first malformed cargo-auditable package before Zod collects an
issue per field. Require cyclic dependency indices to match the cycles in
package dependencies. Keep read-only ELF ranges when a section name does
not fit the bounded name window, so a rustc path in that range is still
reported.
* fix(ci): repair integrated checks and Hopper test baseline
* fix(go): verify Mach-O sections against segments
---------
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Error results reach the wire as text plus `isError: true`, but the
text-only envelope omitted the flag. An error up to 15 bytes above the
result budget passed both the delivery and transport checks and was sent
as the original diagnostic instead of taking the oversized-error recovery
path.
Rename the error representation to "error-text" and size its envelope
with the flag, so ToolResultDelivery and the error transport share the
complete delivered shape. Successful structured results are unchanged.
Fixes#2557
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* perf(evidence): reuse sealed bundle reads in compare and export
`rea compare`, the compare_bundles MCP tool and `rea evidence-export`
read bundles through the mutable reader, so each consumer
re-authenticated every record again: compareBundles parsed both bundles
a second time and bundleDigest's serializeEvidenceBundle a third time;
export's writer parsed the source once more before serializing.
These consumers only read the bundle, so read it through the existing
sealed reader (readImmutableEvidenceBundle). Later parses hit the
immutable bundle cache. Ordinary mutable reads, kept deliberately by
#1721, are unchanged. Output is byte-identical.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(process): report no remaining time after a provider deadline fires
Node timers can fire while performance.now() is still a fraction of a
millisecond short of the monotonic deadline. ProviderStartupDeadline then
reported interruption "timeout" and an aborted signal while remainingMs()
returned ceil(0.x) = 1, so polling loops guarded by remainingMs() > 0
(Hopper and Ghidra startup) could run past an elapsed deadline, and the
real-timer test "aborts a bounded wait when its real monotonic deadline
expires" failed intermittently in CI (expected 1 to be 0).
Once the deadline's own timer has classified a timeout, report zero
remaining time. A deterministic regression test fires the timer with the
monotonic clock 0.4 ms short of the deadline.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Keep failed stable-read handles and WABT runtime roots under retryable resource scopes. Preserve cancellation causes and primary errors, close the same service instances bound to MCP tools, and consolidate CLI operation/cleanup projection.
Validation: fast typecheck and lint, cached runtime build, docs checks, SDK schema validation, focused service and real-file cleanup tests, CLI cleanup regressions, and MCP lifecycle/shutdown workflows. Provider process fixtures simulate engine output; real WABT and Linux EVM execution are not established by this pass.
* perf(native): add opt-in Mach-O summary delivery and retained pages
inspect_macho returns every import and export inline. For VS Code's
Electron Framework one MCP call carries 2.5 MB of text (about 828k o200k
tokens), and the same again as structuredContent; the MCP response
budget admits it. Agents asking for slice identity, UUID or dependency
counts paid that cost on every first call.
Add the established opt-in detail selection: omitted or complete keeps
the existing complete result and Evidence identity, while summary
retains the complete Evidence and returns slice identity with
per-collection returned/total/exhaustive counts (956 tokens).
inspect_analysis_view kind macho pages the original rows of any
collection or the command provenance, repeating the producer's own
total and exhaustive flag so page exhaustion is never read as absence.
Delivery detail never reaches the native tools or the source Evidence
ID: registerEvidenceTools gains a parametersForEvidence hook, so the
summary's parent is the same Evidence as a complete call.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(mcp): refuse unretainable summary delivery before analysis
A server without Evidence retention (a fixed analysis source) ran the
complete inspect_macho or inspect_artifact analysis for detail summary
and only then refused it in summarizeRetainedAnalysis, paying the very
cost summary delivery exists to avoid. registerEvidenceTools gains a
preflight hook; both registrations check retention there, as the WASM,
EVM, managed, pcap and capture handlers already do before dispatch.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(native): keep whole rows in Mach-O architecture and segment pages
The page item union listed plain-object row arrays, and zod returns the
first union member that parses while dropping unknown keys. Entrypoint
rows declare only file_offset, so every architecture and segment row
(which also carry a numeric file_offset) matched it and lost its name,
CPU fields, permissions and sections. Rows now use exact key sets, and
imports/exports share one member. The MCP test pages every facet of a
real /bin/ls and compares it with the unparsed complete rows.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
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>
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>
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.
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>
* feat(server): frame analysis-target content as untrusted data
Strings, decompiled code, page text, scripts, captures and logs returned
by REA originate from the analysed target, which can embed text aimed at
the agent. Nothing marked that content, so a prompt injection inside a
binary or page could steer the agent toward tools that run code.
- Server instructions now state that target-derived tool output is data,
never instructions, and that any request in it to run, install,
download or send something is a prompt-injection attempt to report.
- Tool contracts declare returnsTargetContent. It is on by default; only
seven tools whose results hold REA session state or caller-supplied
values (binary_session, close_binary, export_evidence_bundle,
record_unknown, update_unknown, inspect_adb_client,
inspect_apktool_client) are audited exceptions, so new tools start
marked.
- Successful results of marked tools carry
_meta.provenance = { untrusted_content: true, source: "analysis_target" }.
content and structuredContent are unchanged, and the response budget
accounts for the fixed marker bytes.
Tests cover the encoding and budget accounting, the audited exception
list against the contract catalog, and an in-memory MCP client that sees
the instructions and the marker on a marked tool but not on
binary_session.
* fix(server): mark failed target-content results as untrusted too
The untrusted provenance marker was attached only on the success branch
of ToolResultDelivery, but failures can carry target-derived text as
well: error messages, captured output, and partial observations such as
the terminal frames a failed capture_process_scenario retains in
partial_observation. A host gating later process or device actions on the
marker would have treated an adversarial error payload as trusted.
Every error path in src/server now carries the failing tool's provenance:
- toErrorToolResult and toRecordedPartialErrorToolResult take the tool's
target-content flag explicitly, so each caller is audited by the type
checker; toCallToolResult, toEvidenceToolResult (recording failure), the
response-budget error that replaces an oversized success, and the
Electron partial-evidence path pass contract.returnsTargetContent.
- Admission failures in runAdmittedToolOperation resolve the flag from the
tool name through the new toolReturnsTargetContent, the same predicate
the contract metadata uses; unknown names stay marked.
- The transport's oversized-error replacement keeps the original's marker
and counts it in its budget check, matching the projected error.
- Errors the MCP SDK builds itself (input validation, thrown handlers) are
marked in the transport for target-content tools only.
Error encoding accounts for the marker bytes exactly as success encoding
does. docs/mcp-contracts.md now states that error results carry the
marker. Tests cover marked and unmarked error results, the
capture_process_scenario partial-observation errors in-process and over
the SDK transport, the recorded partial-evidence path, the budget
replacement, the oversized-error replacement, and SDK validation errors
for a marked and a content-free tool.
* fix(contracts): mark session status, close and client probes as target content
The content-free exception set was audited against successful results
only. Now that failures carry the marker too, every exception was
re-audited against its error paths with one question: can the result
include text produced by an external process, provider or device, or
stored content REA did not author?
Removed from TARGET_CONTENT_FREE_TOOLS:
- close_binary: a failed provider cleanup (GhidraClient cleanup error via
createGhidraDiagnostics) projects the launcher's captured stdout and
stderr, which can contain target-derived log text.
- binary_session: capability status reports runtime unavailability as the
provider error message (ProviderAdapterError and similar), and the
provider candidate and Android availability diagnostics include probe
output such as the apktool stderr excerpt.
- inspect_adb_client: the success result carries the version and
installed path parsed from `adb version` stdout, and a failed probe
keeps a stderr_excerpt.
- inspect_apktool_client: the success result carries the first line of
the launcher's `--version` output, and a failed probe keeps a
stderr_excerpt.
- update_unknown: the new revision carries forward the stored question
and domain, which an imported evidence bundle can supply; list_unknowns
and get_evidence_bundle already return that data marked.
The set now holds only export_evidence_bundle (path, byte and record
counts; failures are filesystem, cancellation, or a fixed integrity
message) and record_unknown (the caller's input plus REA-computed ids
and digests; registry errors project only a reason code).
The conformance test pins the reduced set and asserts each removed tool
is marked. Tests that used binary_session as the unmarked example now use
record_unknown. docs/mcp-contracts.md lists the remaining exceptions and
why the removed tools stay marked.
* fix(contracts): include target-content policy in the catalog identity
createToolCatalog() and CATALOG_IDENTITY.tools projected every
client-visible contract field except returnsTargetContent. Because the
untrusted-content framing left schemas, effects, prompts, CLI names and
the package version unchanged, a same-version MCP process started before
the update digested equal to the new server and binary_session reported
it aligned although it never emits _meta.provenance.
Project the flag as returns_target_content in both the digested tool
catalog and CATALOG_IDENTITY.tools, and export createCatalogDigests so a
boundary test proves that toggling the flag on one tool changes the
tools and combined digests.
Treat missing, partial, unknown, and unresolved call coverage as insufficient for call-set equality. Combine IDA caller/callee coverage without losing either explanation and recheck source coverage when verifying historical comparisons.
Scope generated function-comparison unknowns with optional typed comparison/dimension facts, preserving legacy IDs and revision digests when absent. Keep unrelated dimensions conclusive, explicit dependencies applicable, and ambiguous function identity unknown. A confirmed change for matched functions remains a failure even if calls are incomplete.
Validation: function/reconstruction/residual-unknown and bundle tests, recorded IDA protocol regression, public MCP comparison/verification, and advertised SDK JSON Schema checks. Real IDA provider execution was not verified.
Fixes#2222
Inspect selected fields in existing captures through shared CLI/MCP Evidence.
Page matching frames without changing dissection or two-pass read coverage;
retain failed descriptor/process cleanup ownership and reject excess admission.
Verified with real TShark 4.0.17 and 4.2.2, public Wireshark recordings,
owned HTTP application traffic, real CLI/MCP workflows and schema checks.
Co-authored-by: Olorinm <59321404+Olorinm@users.noreply.github.com>
Assisted-by: Kogasa (gensokyo-skills:kogasa-surprise-testing)
Add opt-in EVM MCP summary delivery and retained/portable function and text views. Keep existing complete defaults, provider arguments and original full Evidence identities and records usable. Require retention before selected analysis; preserve inferred candidates, nulls, diagnostics, provenance and explicit unknowns.
Verify old schemas, historical Evidence migration, 64 focused feature checks, 30 affected MCP checks after latest-main rebase, 59 real EVM CLI/MCP cases and six shared-domain artifact workflows. Repeated SDK delivery is smaller; all real model trials and mixed token/time outcomes, limitation failures and resource qualification gaps remain documented without a uniform performance claim.
Refs #2016
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Add offline SQLite database/WAL inspection through the shared CLI/MCP.
Use compact inventory, selected-table schema, exact ordered columns and
bounded row samples while preserving source identity and lossless values.
Retain completed Evidence through late cancellation and private-root cleanup
failures, bound concurrent inspections, and recheck journal validity after
acquisition. Reuse prepared CI runtime artifacts across supported Node lines.
Co-authored-by: Y0UW1SH <29498139+Y0UW1SH@users.noreply.github.com>
Assisted-by: Kogasa (gensokyo-skills:kogasa-surprise-testing)
* fix(native): retain failed pwntools process and runtime cleanup
Keep failed worker/debugger cleanup owners and their private files until
verified process shutdown and successful directory removal. Retry cleanup
before another inspection and expose shutdown through the shared services,
CLI and MCP server. Serialize acquisition without delaying cancellation of
queued callers, and preserve primary diagnostics when CLI cleanup fails.
The original decoder reproduction left a live owned Node child with its
runtime deleted and no retained cleanup owner; the regression now verifies
retention, admission blocking, retries and eventual process/file cleanup.
Validation: 10 lifecycle/public CLI/MCP tests, 20 service tests, test
discovery 814/814, build:cached and full check including architecture guards.
Actual POSIX process/filesystem boundaries used injected Node workers on
macOS; real Linux pwntools/GDB provider parsing was not exercised.
* test(native): verify retained recorded-crash cleanup lifecycle
* fix(native): finalize diagnostic cleanup before publishing results
* fix(native): retire decoder stages before reading runtime evidence
---------
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
Bind existing EVM CLI/MCP metadata and the legacy service factory through one static descriptor, and register the same selected service that owns cleanup. Preserve original names, schemas, inputs/full defaults, configuration, Evidence and injection identity.
Verified exact SDK catalog/fact parity across six real EVM pairs, 59 actual EVM CLI/MCP cases, six real cross-domain workflows, six actual Codex CLI fact/reference/lifecycle trials and relevant local checks. CI run 38283026233 succeeds for exact head d999d9874b. No token reduction from wiring or fully qualified resource claim; final counters remain unknown. Continues #2016.
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Ysopwnable