Commit Graph

1301 Commits

Author SHA1 Message Date
morluto 93731e17dd
fix(ci): decouple stdio smoke tests from Linux display probes (#1014)
* fix(ci): decouple stdio smoke tests from Linux display probes

* test(ci): check compiled catalog without an initial target
2026-10-08 02:50:08 +08:00
Ysopwnable 6a278a68dd
fix(release): support default credentials and record 5.0.0 boundary (#1012)
fix(release): support built-in credentials and refresh release boundary

GitHub now creates approval-required PR workflows for GITHUB_TOKEN-authored
opened/synchronize/reopened events. Restore optional custom credentials and
cover prepare/publish without a separate secret. Record the verified 5.0.0
checkpoint, public 133-tool catalog and package update in the installation guide.

Assisted-by: Sakuya (gensokyo-skills:sakuya-checkpointed-execution)
2026-10-08 02:26:29 +08:00
morluto 76e7e510bb
perf(artifacts): replace quadratic ZIP overlap checks (#1013) 2026-10-08 02:08:39 +08:00
morluto 966cca8efe fix(release): sync 5.0.0 checkpoint back to main 2026-10-08 01:43:02 +08:00
morluto 1d62e54e1f
fix(snapshot): replay validated workflow evidence (#984)
* fix(snapshot): bind replay to validated cache entry

* fix(snapshot): persist composed workflow results

* fix(snapshot): enforce workflow replay identity

* fix(snapshot): honor cancellation after route resolution

* test(snapshot): make cancellation race deterministic

* fix(snapshot): replay all composed workflow evidence

* fix(snapshot): index workflow evidence once per snapshot

Validate workflow entries against a query and content index instead of scanning the evidence bundle for every entry.

* fix(snapshot): reuse previewed routes and bound workflow cache

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
2026-10-08 01:33:39 +08:00
github-actions[bot] b33236ec08
chore(release/5.0.0): release rea-agents 5.0.0 (#1007)
* chore(release/5.0.0): release rea-agents 5.0.0

* chore: normalize release product catalog

---------

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
2026-10-08 01:26:36 +08:00
DLME2024 96a46ada0a
fix(javascript): bound semantic node projection for resource safety (#966)
* fix(javascript): bound semantic node projection for resource safety

analyze_javascript_application exhausted the heap or failed serialization on
ordinary inputs. A small extracted tree crashed outright, and a larger
application tree failed at both the default heap and an 8 GB heap.

Cause: the semantic projection emits one node per AST expression, call site,
binding and property slot, with no ceiling. A five-file tree already yields on
the order of 170,000 nodes, and the whole result is serialized after the object
graph is built and validated. Result ceilings were deliberately removed upstream,
so nothing bounded this growth.

Apply two resource-safety ceilings and report them through the graph's own
coverage contract rather than dropping data silently:

- SEMANTIC_GRAPH_NODE_CEILING bounds retained nodes per tree (100,000, roughly
  300 MB serialized, below V8's 512 MB string limit).
- SEMANTIC_GRAPH_FILE_NODE_CEILING bounds each source file (20,000) so one
  bundled library cannot starve the application's own modules.

Exhausting either marks coverage truncated with a partial status and records the
ceiling in coverage.limits. A small application stays complete and untruncated.

Verified: a previously-crashing application tree now returns Evidence in under a
minute, while a small tree stays complete. Existing tests and check:fast pass.

* fix(javascript): report semantic graph truncation accurately

---------

Co-authored-by: REA Contributor <contributor@users.noreply.github.com>
Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-08 01:22:10 +08:00
morluto b4fc0ecaa1
fix(release): publish from explicit source checkpoints (#1006)
* fix(release): publish from explicit source checkpoints

Dispatch preparation and publication against a frozen release branch.
Bind package builds and provenance to the same tagged commit, enable
release-branch PR CI, and document the maintainer workflow.

Assisted-by: Sakuya (gensokyo-skills:sakuya-checkpointed-execution)

* fix(release): stop publication when the checkpoint has moved

Reject unknown phases and a release-branch tip that differs from the
dispatch SHA before Release Please creates a tag. Keep registry jobs and
catalog pushes on their own phase, and run website checks for release PRs.

* fix(release): require a token that can trigger release CI

Stop falling back to GITHUB_TOKEN, which cannot start workflow runs for the release pull request it creates.

* test(release): require the CI-capable release token

Lock the workflow to RELEASE_PLEASE_TOKEN and fail preparation when that secret is missing.

* style(release): format the release-token test

---------

Co-authored-by: N0zoM1z0 <1705004398@qq.com>
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
2026-10-08 01:15:54 +08:00
Ysopwnable 64c71880ae
fix(release): publish from explicit source checkpoints (#1003)
* fix(release): publish from explicit source checkpoints

Dispatch preparation and publication against a frozen release branch.
Bind package builds and provenance to the same tagged commit, enable
release-branch PR CI, and document the maintainer workflow.

Assisted-by: Sakuya (gensokyo-skills:sakuya-checkpointed-execution)

* fix(release): drive checkpoint releases from frozen branches

Use branch pushes for Release Please preparation and reviewed-merge publication.
Reject moved source refs before tagging and bind registry provenance to the
triggering release commit. Remove phase inputs and document the three steps.

Assisted-by: Sakuya (gensokyo-skills:sakuya-checkpointed-execution)
Assisted-by: Cirno (gensokyo-skills:cirno-radical-simplification)
2026-10-08 01:12:18 +08:00
morluto fe64fd65d9
fix(mcp): reject unknown MCP tool inputs (#975)
* fix(mcp): reject unknown official tool inputs

* fix(mcp): reject unknown native tool inputs

* test(mcp): follow native contract module move

* fix(mcp): reject unknown remaining tool inputs

* fix(mcp): satisfy strict unknown and browser input schemas

Remove obsolete approval fields from residual-unknown fixtures, project update inputs explicitly, and stop passing target_id to list-targets parsing.

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
2026-10-08 01:10:53 +08:00
morluto ae0f725fad
fix(browser): preserve authorized redirect hops (#986)
* fix(browser): preserve authorized redirect hops

* fix(browser): compare redirect semantics without event times

* fix(browser): classify missing redirect predecessors

* fix(browser): distinguish malformed redirect URLs

* fix(browser): preserve evidence on malformed redirect envelopes

* fix(browser): freeze evidence after malformed redirect

* docs: regenerate catalogs after main integration

* fix(browser): account for skipped redirect body shapes

* fix(browser): mark skipped redirect response shapes unavailable

When a malformed redirect keeps a JSON predecessor, record that its response body shape was not captured.

---------

Co-authored-by: Cursor Agent <cursoragent@cursor.com>
2026-10-08 01:08:31 +08:00
morluto a6132be054
fix(cli): classify ENOTDIR JSON input failures (#1000)
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: morluto <morluto@users.noreply.github.com>
2026-10-08 01:06:31 +08:00
N0zoM1z0 fc58fa1eb0 fix(release): drive checkpoint releases from frozen branches
Use branch pushes for Release Please preparation and reviewed-merge publication.
Reject moved source refs before tagging and bind registry provenance to the
triggering release commit. Remove phase inputs and document the three steps.

Assisted-by: Sakuya (gensokyo-skills:sakuya-checkpointed-execution)
Assisted-by: Cirno (gensokyo-skills:cirno-radical-simplification)
2026-10-07 17:02:03 +00:00
morluto cbc0dd1e7c
fix(contracts)!: require absolute local paths for remaining caller-supplied file inputs (#1004)
* fix(contracts)!: require absolute local paths for remaining caller-supplied file inputs

Follow-up to the session snapshot/export tightening: the same invisible server-cwd resolution applied to open_binary target paths (parseBinaryTarget(path, process.cwd())), import_evidence_bundle bundle reads, firmware input files (resolve(raw)), browser executable launch paths (resolve-then-spawn), Electron executable/application/root paths (realpath on raw input), and the expected server path check.

Each field gains the shared isAbsoluteLocalPath refinement plus a catalog-visible description, since JSON Schema cannot express the rule. Examples, scenario fixtures, and contract tests updated to absolute paths; new schema cases cover rejection/acceptance/omission per field.

Explicitly unchanged (verified rooted or out of scope): inspect_plist paths (resolved against the target bundle dir), firmware output_directory (already provider-enforced), filesystem observation paths (resolved against the caller-supplied working_directory), keyed-archive member paths (archive-relative), JS evidence member paths (evidence-relative under root_alias), provider-internal execution schemas, CLI-only doctor flows, and CLI options (own resolve against the operator cwd).

* fix(cli): resolve operator-relative paths before shared absolute-path validation

The contract tightening rejects relative paths, but three CLI commands forward operator-relative input into the shared schemas: firmware commands pass raw args to FirmwareAnalysisService, and the Electron/browser scenario commands validate parsed scenario JSON directly. CLI paths are relative to the operator shell, so resolve them against the working directory first: resolve() in firmwareCommands, and a shared resolveCliJsonPaths JSON key-path helper (cliJsonInput) applied in electronCommands and cliBrowserScenarioCommands. MCP behavior is unchanged — MCP callers must still send absolute paths.

Covers the regression with a relative-path firmware CLI acceptance test (asserts provider-stage capability_unavailable, not a validation error) and unit tests for the helper.
2026-10-08 00:52:37 +08:00
Ysopwnable eb8f6491a8
fix(ci): generate source catalog before shard tests (#1002) 2026-10-08 00:47:11 +08:00
Ysopwnable 6aa9afb549
refactor(process): separate capture implementation from analyst workflows (#988)
* refactor(process): separate capture implementation from analyst workflows

* fix(ci): generate source catalog before shard tests
2026-10-08 00:46:21 +08:00
N0zoM1z0 49887f0239 fix(release): publish from explicit source checkpoints
Dispatch preparation and publication against a frozen release branch.
Bind package builds and provenance to the same tagged commit, enable
release-branch PR CI, and document the maintainer workflow.

Assisted-by: Sakuya (gensokyo-skills:sakuya-checkpointed-execution)
2026-10-07 16:38:02 +00:00
Dery Ferdika 5b172b1042
fix(contracts)!: require absolute local paths for session filesystem inputs (#948)
fix(contracts): require absolute local paths for session filesystem inputs

Snapshot, evidence-bundle export, and managed-target inputs accepted any non-empty string, so relative paths were resolved against the MCP server working directory instead of the caller's intent. Refine these contract fields to absolute host paths (Windows drive and forward-slash forms included), advertise the constraint on the property descriptions, and regenerate the derived catalog, product catalog, and skill catalog digest. Relative CLI options are unaffected because CLI flows do not parse these MCP contracts.

Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-08 00:31:55 +08:00
Yunare Maia 06b047a092
fix(release): extend npm propagation window and improve diagnostics (#943)
* fix(release): extend npm propagation window and improve diagnostics

The release verification step polls npm for up to 20 minutes (80 attempts
× 15s), but npm registry propagation can take longer. When the version
doesn't appear within the window, the workflow fails even though the
package was already published successfully.

Changes:
- Extend polling window from 20 to 30 minutes (120 attempts × 15s)
- Add --force flag to npm view to bypass cache on each poll
- Report full retry window in diagnostics (attempts × sleep time)
- Distinguish propagation delay from CLI verification failure in error message

Fixes #621

* fix(release): drop no-op --force from npm propagation polling

npm view --force is accepted but only prints a protections warning on a read-only command; --prefer-online already forces fresh registry metadata on each poll.

---------

Co-authored-by: morluto <76467478+morluto@users.noreply.github.com>
2026-10-08 00:26:25 +08:00
Ysopwnable 161ed91571
test(process): wait for observable snapshot ctime changes (#999) 2026-10-08 00:24:44 +08:00
morluto 6f0ef16050
fix(build): clarify generated-file check hint and drop redundant catalog step (#998)
The shared stale-file hint told developers to commit regenerated output, which is wrong for the gitignored build catalog. mcp-catalog:generate no longer re-runs the generator after build:cached since build:unlocked already produces it.
2026-10-08 00:21:39 +08:00
morluto bda4ada7fb
fix(dotnet): validate managed metadata and CIL boundaries (#938)
* perf(dotnet): reuse decoded metadata rows

* test(dotnet): preserve expanded fixture metadata

* style(dotnet): format metadata inventory

* fix(dotnet): bound method bodies to PE section data

* fix(dotnet): derive fixture metadata index widths

* fix(dotnet): validate managed metadata bounds

* test(dotnet): cover large metadata ReadyToRun layout

* fix(dotnet): validate exception instruction boundaries

* fix(dotnet): report invalid metadata coded indexes

* fix(dotnet): clarify MemberForwarded null diagnostics

* fix(dotnet): honor complete CIL and EH boundaries

* fix(dotnet): validate metadata identity rows and offsets
2026-10-08 00:17:26 +08:00
morluto 35a4e18d30
refactor(build): generate MCP tool catalog at build time instead of committing it (#997)
src/generatedMcpToolCatalog.ts had zero runtime importers and churned on every contract edit, guaranteeing merge conflicts on long-lived branches. It is now gitignored and regenerated by build:unlocked; typecheck and knip depend on build via turbo. Tests consume the build output unchanged. Adds a regen-generated CI workflow that pushes tracked generated metadata (product catalog, skill frontmatter, verification manifest/ledger, package metadata, error schema) to same-repo PR branches.
2026-10-08 00:16:34 +08:00
Kaoru0822 c67530535a
Merge pull request #994 from HoneyTyagii/perf/canonical-digest-batching-904
perf(domain): batch canonical digest writes with bounded buffer
2026-10-07 09:14:23 -07:00
Kaoru0822 701efd5f90
Merge pull request #996 from morluto/docs/ja-readme-polish
docs: polish the Japanese README and align it with current facts
2026-10-07 09:12:10 -07:00
Kaoru0822 e45da24297 docs: polish the Japanese README and align it with current facts
Correct statements that drifted from the English README: IDA Pro was
missing, the diagram labeled Ghidra read-only although Linux/macOS
support annotations, Android verification named Linux instead of macOS
arm64, and the security model described only Unix sockets. Add scoped
doctor checks, provider selection, the Ghidra/Java FAQ, and Hopper
request behavior, and polish phrasing and terminology throughout.

Refs #958

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 16:03:37 +00:00
morluto 4a6073810a
fix(contracts): align CLI/MCP boundary contracts (#946)
* fix(cli): validate native scenario and preserve JSON paths

* fix(process): reject invalid UTF-8 JSON inputs

* fix(cli): retain missing JSON path diagnostics

* test(cli): cover relative brace and bracket JSON paths

* test(cli): reject missing native scenario steps

* fix(mcp): advertise reserved process environment key

* fix(process): reject NUL in process scenario values

* fix(process): advertise NUL input constraints to MCP

* fix(cli): report scenario input schema issues

* fix(contracts): align CLI and MCP input validation

* fix(mcp): document semantic output constraints

* fix(mcp): describe process journal compatibility

* docs: regenerate catalog after schema grouping

* docs: regenerate catalog after main merge

* fix(cli): preserve JSON and native UI input diagnostics

* test(mcp): reuse advertised tool schemas in managed flow
2026-10-08 00:03:36 +08:00
morluto fc4a9a0592
fix(artifacts): preserve reader failure diagnostics (#987)
* fix(artifacts): preserve filesystem and DMG diagnostics

* fix(artifacts): classify DMG verification failures

* fix(artifacts): distinguish inaccessible disk images

* fix(artifacts): classify damaged image checksums
2026-10-07 23:58:07 +08:00
morluto abd7a5ed09
perf(artifacts): stream ASAR members without full buffering (#962)
* fix(artifacts): stream ASAR members without full buffering

* fix(artifacts): close unopened ASAR member streams

* fix(artifacts): preserve unpacked ASAR mismatch digests

* fix(artifacts): preserve ASAR read failures

* fix(artifacts): retain unpacked ASAR stat failures

* fix(artifacts): validate ASAR entry producer identity
2026-10-07 23:57:05 +08:00
morluto 58381eefa2
perf(inspector): bound runtime script location authorization (#945)
* perf(inspector): authorize script locations concurrently

* fix(inspector): bound concurrent location authorization

* fix(inspector): stop queued authorization on failure

* test(inspector): avoid freezing authorization scheduling
2026-10-07 23:54:31 +08:00
morluto ebb065e2dc
perf(comparison): reuse canonical sort keys (#990) 2026-10-07 23:54:09 +08:00
morluto 2cedde9f1c
perf(javascript): index Promise return-site ownership (#971)
* perf(javascript): index promise return ownership lookups

* perf(javascript): defer Promise return-site indexing
2026-10-07 23:53:41 +08:00
morluto 248972f327
fix(installer): validate exact semantic versions (#954)
* fix(installer): validate exact semantic versions

* fix(installer): reject npm-ambiguous build metadata
2026-10-07 23:53:26 +08:00
Ysopwnable 5047373006
feat(website): add a shared back-to-top link
Long guides and case studies now have a small `↑ Top` link at the bottom right. It appears after scrolling, returns smoothly to the page header, and moves keyboard focus back to navigation. Reduced-motion users return immediately; the native `#top` link also works without JavaScript.

The link is shared by all 11 styled website pages. The README and style guide document it, and the evidence notes retain the original browser-example digest alongside its later navigation addition.

Validation:

- Website preparation and verification: 12 HTML pages, 195 local references, SVG XML, the six-file ZIP and the sole manual Pages publisher.
- Oxfmt, `git diff --check` and `npm run check:fast` with Node.js 24.18.0.
- Focused browser checks at `/` and `/rea/`: 22 routes, 21 pointer returns, six keyboard returns, five viewport widths, smooth scrolling, reduced motion, focus return and the no-JavaScript anchor.
- Existing site interactions: 24 routes, 68 copy buttons, 20 comparison steps, two CSV exports, two ZIP downloads and eight viewport widths from 320px to 1440px. Code excerpts are unchanged.
- Desktop and mobile rendering reviewed. Downloadable Electron source and CSV export source/data are unchanged.

Publication continues through the existing manual Pages workflow after merge.
2026-10-07 23:50:31 +08:00
morluto 7dc40d3d9e
perf(evidence): scale imports and dependency graphs (#941)
* perf(evidence): avoid rebuilding ledger on record imports

* perf(evidence): scale unknown graph validation

* fix(evidence): preserve import and projection semantics

* fix(evidence): invalidate cache on metadata replacement

* test(evidence): isolate snapshot import regression

* fix(evidence): preserve snapshots on additive imports
2026-10-07 23:47:47 +08:00
Honey Tyagi dca14900c4 perf(domain): batch canonical digest writes with bounded buffer 2026-10-07 21:16:44 +05:30
morluto 93aa9eb5c2
fix(process): harden filesystem snapshot identity and cancellation (#956)
* fix(process): bind snapshots to captured file identity

* fix(process): honor snapshot cancellation for empty files

* test(process): verify snapshot descriptor cleanup

* fix(process): preserve snapshot cancellation results

* fix(process): preserve non-cancellation capture failures

* test(process): verify descriptor cleanup with fstat

* fix(process): detect snapshot changes by ctime

* test(process): prove snapshot ctime invalidation
2026-10-07 23:45:23 +08:00
morluto d76f7d1508
fix(windows): remove cancelled snapshot outputs (#957)
* fix(windows): remove cancelled snapshot outputs

* fix(windows): allow readback of retained runtime files

* test(windows): verify retained runtime file protection
2026-10-07 23:45:16 +08:00
morluto 9f496ef2d4
fix(native)!: preserve incomplete native UI captures (#991)
* fix(native): preserve unknown accessibility child counts

* fix(native): align helper buffer with capture budget

* fix(native): align Swift helper JSON wire size

* fix(native): bound raw helper output to scenario budget
2026-10-07 23:44:41 +08:00
Ysopwnable 5065fcd520
docs(website): clarify guides and case studies
Several guides use broad headings and repeat their summaries, making the main example harder to follow. This change names each task directly, keeps Notion's clipboard investigation in the main reading path, and moves the Markdown example and longer provenance into expandable details. The browser guide now opens the hosted example, and the first native CLI query works without a global install.

The existing figures, prompts, source excerpts and comparison controls remain the teaching material. Provider details now fit narrow screens. The repository README links to all Showcases, and the website authoring docs share a signed design principle: a good website is like a good paper—easy to follow, clear and concise, with a clean, refined presentation. A new style guide and scoped AGENTS.md explain how to maintain that approach.

Closes #985.

Validation:

- `python3 scripts/prepare-website.py` and `python3 scripts/verify-website.py`: 12 HTML pages, 183 local references, SVG XML, the six-file ZIP and the sole manual Pages publisher.
- Oxfmt checks and `git diff --check`; `npm run check:fast` with Node.js 24.18.0.
- Browser checks at both `/` and `/rea/`: 24 routes, 68 copy buttons, 20 comparison steps, two CSV exports, two ZIP downloads and eight viewport widths from 320px to 1440px.
- Existing code excerpts, expanded details, keyboard access, no-JavaScript content, continuous cursor animation and reduced motion checked.

Existing analysis and compiler-check evidence is retained; real native-provider verification and compiler replay were not rerun for these copy and layout changes. The homepage content and publication triggers are unchanged.
2026-10-07 23:24:15 +08:00
morluto 0a4caf4a0a
fix(evaluation): canonicalize Unicode argument keys (#980) 2026-10-07 23:19:49 +08:00
Namnarak 98b2b027af
fix(linux): improve Hopper reliability and CachyOS support (#963)
* fix(linux): support CachyOS Hopper hosts

* fix(config): reuse user-local Hopper launcher

* fix(hopper): harden Linux demo dialog handling

* chore: retrigger CI after PR title fix

* fix(linux): align Hopper defaults and support docs

---------

Co-authored-by: Ysopwnable <161784452+N0zoM1z0@users.noreply.github.com>
2026-10-07 23:16:48 +08:00
morluto 8aaf9a1978
fix(firmware): retain multi-file extraction failures (#978) 2026-10-07 23:06:41 +08:00
morluto d5c43d65f0
fix(cli): preserve output and comparison input validation (#981) 2026-10-07 23:04:54 +08:00
morluto a45fe86a39
fix(build): validate installed dependencies before cache hits (#982) 2026-10-07 23:01:01 +08:00
morluto 71ac83f3dd
fix(apple): traverse keyed archive hierarchy collections (#976) 2026-10-07 23:00:43 +08:00
morluto fb3749aff6
fix(android): align runtime suffix inference (#977) 2026-10-07 23:00:22 +08:00
morluto f26c8b5b8c
perf(reconstruction): index coverage evaluation (#983) 2026-10-07 22:59:13 +08:00
Ysopwnable ea4db4c9f7
refactor(native): group analyst semantics workflows and contracts (#973) 2026-10-07 22:40:48 +08:00
Ysopwnable f8d84da127
docs(website): add TH04 bullet-ring case study
Adds a TH04 case study showing how its original 16-bit PC-98 DOS code spaces a bullet ring and aims it at the player. Fresh REA instructions are paired with a five-step C++ comparison and a ring diagram; homepage and showcase-index entries link to the new page and the requested reconstruction repository.

The page credits TH04's maintained source and recorded compiler checks separately from the new REA inspection. Public provenance pins the target, source checkpoint and evidence. No original game assets, executables, private paths or raw result bundles are included. Pages remains manually dispatched.

Validation:
- Website verifier: 12 HTML pages, 185 local references, SVG XML and downloadable ZIP.
- Browser: root and `/rea/`, 32 copy checks, five comparison steps, eight viewport widths, keyboard, reduced motion and no-JavaScript content.
- Fresh real Ghidra verification through published REA 4.1.0: DOS load image verified, all 84 checks match; shown instructions match the function dossier. No Hopper verification was needed or run.
- Independent claims/privacy reviews passed; seven immutable public source links match the local TH04 checkpoint.
- `npm run check:fast`.

The two clean compiler replays and 2,139 matching bytes are recorded TH04 checks, reviewed against their retained receipt and source hashes; the website update does not rerun the compiler or game.
2026-10-07 22:09:16 +08:00