fix(cv): harden document parsing

Upgrade and hash-lock upload-facing parser dependencies, reject resource-heavy or mismatched inputs, remove unsafe backend binary fallbacks, and prevent internal parser failures from leaking to users.

JobTrackerApi.Tests/CvExtractionCoverageTests.cs

@@ -1,5 +1,6 @@
 using System.Reflection;
 using System.IO.Compression;
+using System.Text;
 using JobTrackerApi.Controllers;
 using JobTrackerApi.Models;
 using Microsoft.AspNetCore.Http;
@@ -241,7 +242,7 @@ public sealed class CvExtractionCoverageTests
     }
 
     [Fact]
-    public async Task Docx_fallback_preserves_paragraph_and_table_boundaries()
+    public async Task Binary_document_fallback_is_disabled()
     {
         await using var package = new MemoryStream();
         using (var archive = new ZipArchive(package, ZipArchiveMode.Create, leaveOpen: true))
@@ -263,14 +264,23 @@ public sealed class CvExtractionCoverageTests
         }
         package.Position = 0;
         var upload = new FormFile(package, 0, package.Length, "file", "sanitized.docx");
-        var method = typeof(ProfileCvController).GetMethod("ExtractTextAsync", BindingFlags.NonPublic | BindingFlags.Static)!;
+        var method = typeof(ProfileCvController).GetMethod("ExtractPlainTextAsync", BindingFlags.NonPublic | BindingFlags.Static)!;
-        var task = Assert.IsAssignableFrom<Task<string>>(method.Invoke(null, new object[] { upload, ".docx" }));
+        var task = Assert.IsAssignableFrom<Task<string>>(method.Invoke(null, new object[] { upload, ".docx", CancellationToken.None }));
         var extracted = await task;
 
-        Assert.Contains("Technical Skills\n", extracted);
+        Assert.Empty(extracted);
-        Assert.Contains("Backend | C#, .NET", extracted);
+    }
-        Assert.Contains("DevOps | Docker, Linux", extracted);
+
-        Assert.Contains("\n\nExample Engineer\n- Built reliable services", extracted);
+    [Fact]
+    public async Task Plain_text_fallback_reads_bounded_utf8()
+    {
+        const string source = "# Ada Lovelace\n\n## Skills\nC#\nSQL";
+        await using var stream = new MemoryStream(Encoding.UTF8.GetBytes(source));
+        var upload = new FormFile(stream, 0, stream.Length, "file", "resume.md");
+        var method = typeof(ProfileCvController).GetMethod("ExtractPlainTextAsync", BindingFlags.NonPublic | BindingFlags.Static)!;
+        var task = Assert.IsAssignableFrom<Task<string>>(method.Invoke(null, new object[] { upload, ".md", CancellationToken.None }));
+
+        Assert.Equal(source, await task);
     }
 
     private static StructuredCvProfile InvokeProfileBuilder(string markdown)

JobTrackerApi.Tests/ProfileCvControllerTests.cs

@@ -1112,6 +1112,70 @@ public sealed class ProfileCvControllerTests
         Assert.Equal("pending_review", run.Status);
     }
 
+    [Fact]
+    public async Task Binary_upload_does_not_fall_back_to_in_process_parsing()
+    {
+        var user = new ApplicationUser { Id = "user-1" };
+        var userManager = CreateUserManager();
+        userManager.Setup(x => x.GetUserAsync(It.IsAny<ClaimsPrincipal>())).ReturnsAsync(user);
+        ConfigureWorkerUser(userManager, user);
+        userManager.Setup(x => x.UpdateAsync(user)).ReturnsAsync(IdentityResult.Success);
+        var aiService = new Mock<ISummarizerService>();
+        aiService
+            .Setup(x => x.ExtractTextAsync(It.IsAny<Stream>(), It.IsAny<string>(), It.IsAny<string?>(), It.IsAny<CancellationToken>()))
+            .ReturnsAsync((AiTextExtractionResult?)null);
+
+        await using var db = CreateDb();
+        var controller = CreateController(userManager.Object, aiService.Object, db, CreatePaths());
+        var bytes = Encoding.UTF8.GetBytes("not a safe DOCX package");
+        var file = new FormFile(new MemoryStream(bytes), 0, bytes.Length, "file", "resume.docx")
+        {
+            Headers = new HeaderDictionary(),
+            ContentType = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+        };
+
+        Assert.IsType<AcceptedResult>(await controller.Upload(file));
+        var run = await db.CvExtractionRuns.SingleAsync();
+        var outcome = Assert.IsType<CvProcessingOutcome>(await controller.ProcessQueuedRunAsync(run.Id, CancellationToken.None));
+
+        Assert.False(outcome.Succeeded);
+        Assert.Equal("The document extraction service could not read this CV safely.", outcome.FailureMessage);
+        Assert.Equal("failed", run.Status);
+        Assert.Equal(outcome.FailureMessage, run.ErrorMessage);
+    }
+
+    [Fact]
+    public async Task Unexpected_extraction_errors_are_not_exposed_to_users()
+    {
+        var user = new ApplicationUser { Id = "user-1" };
+        var userManager = CreateUserManager();
+        userManager.Setup(x => x.GetUserAsync(It.IsAny<ClaimsPrincipal>())).ReturnsAsync(user);
+        ConfigureWorkerUser(userManager, user);
+        userManager.Setup(x => x.UpdateAsync(user)).ReturnsAsync(IdentityResult.Success);
+        var aiService = new Mock<ISummarizerService>();
+        aiService
+            .Setup(x => x.ExtractTextAsync(It.IsAny<Stream>(), It.IsAny<string>(), It.IsAny<string?>(), It.IsAny<CancellationToken>()))
+            .ThrowsAsync(new IOException(@"Secret parser path C:\private\cv.docx"));
+
+        await using var db = CreateDb();
+        var controller = CreateController(userManager.Object, aiService.Object, db, CreatePaths());
+        var source = Encoding.UTF8.GetBytes("# Resume");
+        var file = new FormFile(new MemoryStream(source), 0, source.Length, "file", "resume.md")
+        {
+            Headers = new HeaderDictionary(),
+            ContentType = "text/markdown"
+        };
+
+        Assert.IsType<AcceptedResult>(await controller.Upload(file));
+        var run = await db.CvExtractionRuns.SingleAsync();
+        var outcome = Assert.IsType<CvProcessingOutcome>(await controller.ProcessQueuedRunAsync(run.Id, CancellationToken.None));
+
+        Assert.False(outcome.Succeeded);
+        Assert.Equal("CV processing failed unexpectedly. Please try again.", outcome.FailureMessage);
+        Assert.Equal(outcome.FailureMessage, run.ErrorMessage);
+        Assert.DoesNotContain("private", outcome.FailureMessage, StringComparison.OrdinalIgnoreCase);
+    }
+
     [Fact]
     public void Normalized_markdown_parse_preserves_real_estate_job_and_language_levels()
     {

JobTrackerApi/Controllers/ProfileCvController.Parsing.cs

@@ -3,7 +3,6 @@ using System.Security.Cryptography;
 using System.Text;
 using System.Text.Json;
 using System.Text.RegularExpressions;
-using System.Xml.Linq;
 using JobTrackerApi.Data;
 using JobTrackerApi.Services;
 using JobTrackerApi.Models;
@@ -1352,78 +1351,29 @@ public sealed partial class ProfileCvController : ControllerBase
             || Regex.IsMatch(text, @"(?im)^\s*#\s*(Contact|Professional Summary|Summary|Work Experience|Experience|Education|Skills|Languages|Interests)");
     }
 
-    private static async Task<string> ExtractTextAsync(IFormFile file, string extension)
+    private static async Task<string> ExtractPlainTextAsync(IFormFile file, string extension, CancellationToken cancellationToken)
     {
-        if (string.Equals(extension, ".txt", StringComparison.OrdinalIgnoreCase) || string.Equals(extension, ".md", StringComparison.OrdinalIgnoreCase))
+        if (!string.Equals(extension, ".txt", StringComparison.OrdinalIgnoreCase) &&
+            !string.Equals(extension, ".md", StringComparison.OrdinalIgnoreCase)) return string.Empty;
+
+        using var stream = file.OpenReadStream();
+        using var reader = new StreamReader(
+            stream,
+            new UTF8Encoding(encoderShouldEmitUTF8Identifier: false, throwOnInvalidBytes: true),
+            detectEncodingFromByteOrderMarks: true);
+        var buffer = new char[8192];
+        var result = new StringBuilder(capacity: (int)Math.Min(file.Length, 64 * 1024));
+        while (true)
         {
-            using var stream = file.OpenReadStream();
+            var read = await reader.ReadAsync(buffer.AsMemory(), cancellationToken);
-            using var reader = new StreamReader(stream, Encoding.UTF8, detectEncodingFromByteOrderMarks: true);
+            if (read == 0) break;
-            return (await reader.ReadToEndAsync()).Trim();
+            if (result.Length + read > 200_000)
-        }
-
-        await using var memory = new MemoryStream();
-        await file.CopyToAsync(memory);
-        var bytes = memory.ToArray();
-
-        if (string.Equals(extension, ".pdf", StringComparison.OrdinalIgnoreCase))
-        {
-            var raw = Encoding.Latin1.GetString(bytes);
-            var textMatches = Regex.Matches(raw, @"\((.*?)\)Tj", RegexOptions.Singleline)
-                .Select(match => match.Groups[1].Value)
-                .Concat(Regex.Matches(raw, @"\[(.*?)\]TJ", RegexOptions.Singleline)
-                    .SelectMany(match => Regex.Matches(match.Groups[1].Value, @"\((.*?)\)", RegexOptions.Singleline).Select(x => x.Groups[1].Value)))
-                .Where(value => !string.IsNullOrWhiteSpace(value))
-                .Select(value => Regex.Unescape(value))
-                .ToList();
-
-            var joined = textMatches.Count > 0 ? string.Join(" ", textMatches) : raw;
-            var scrubbed = Regex.Replace(joined, @"[\x00-\x08\x0B\x0C\x0E-\x1F]", " ");
-            return Regex.Replace(scrubbed, @"\s+", " ").Trim();
-        }
-
-        if (string.Equals(extension, ".docx", StringComparison.OrdinalIgnoreCase))
-        {
-            using var archive = new System.IO.Compression.ZipArchive(new MemoryStream(bytes), System.IO.Compression.ZipArchiveMode.Read, leaveOpen: false);
-            var entry = archive.GetEntry("word/document.xml");
-            if (entry is null) return string.Empty;
-            using var entryStream = entry.Open();
-            using var reader = new StreamReader(entryStream, Encoding.UTF8);
-            var xml = await reader.ReadToEndAsync();
-            var document = XDocument.Parse(xml, LoadOptions.PreserveWhitespace);
-            XNamespace word = "http://schemas.openxmlformats.org/wordprocessingml/2006/main";
-            var body = document.Root?.Element(word + "body");
-            if (body is null) return string.Empty;
-
-            static string Text(XElement element, XNamespace ns) => string.Concat(
-                element.Descendants(ns + "t").Select(node => node.Value));
-
-            var blocks = new List<string>();
-            foreach (var block in body.Elements())
             {
-                if (block.Name == word + "p")
+                throw new InvalidOperationException("The extracted CV text is too large to process safely.");
-                {
-                    var paragraph = Text(block, word).Trim();
-                    if (paragraph.Length == 0) continue;
-                    var style = block.Element(word + "pPr")?.Element(word + "pStyle")?.Attribute(word + "val")?.Value ?? string.Empty;
-                    if (style.Contains("Role", StringComparison.OrdinalIgnoreCase) && blocks.Count > 0) blocks.Add(string.Empty);
-                    blocks.Add(style.Contains("Bullet", StringComparison.OrdinalIgnoreCase) ? $"- {paragraph}" : paragraph);
-                    continue;
-                }
-
-                if (block.Name != word + "tbl") continue;
-                foreach (var row in block.Elements(word + "tr"))
-                {
-                    var cells = row.Elements(word + "tc")
-                        .Select(cell => string.Join(" ", cell.Elements(word + "p").Select(paragraph => Text(paragraph, word).Trim()).Where(value => value.Length > 0)))
-                        .Where(value => value.Length > 0)
-                        .ToList();
-                    if (cells.Count > 0) blocks.Add(string.Join(" | ", cells));
-                }
             }
-
+            result.Append(buffer, 0, read);
-            return string.Join("\n", blocks).Trim();
         }
 
-        return string.Empty;
+        return result.ToString().Trim();
     }
 }

JobTrackerApi/Controllers/ProfileCvController.Pipeline.cs

@@ -231,11 +231,15 @@ public sealed partial class ProfileCvController : ControllerBase
 
         if (string.IsNullOrWhiteSpace(text))
         {
-            text = (await ExtractTextAsync(file, extension)).Trim();
+            text = (await ExtractPlainTextAsync(file, extension, cancellationToken)).Trim();
         }
         if (string.IsNullOrWhiteSpace(text))
         {
-            throw new InvalidOperationException("The uploaded CV file could not be read or was empty.");
+            throw new InvalidOperationException(
+                string.Equals(extension, ".txt", StringComparison.OrdinalIgnoreCase) ||
+                string.Equals(extension, ".md", StringComparison.OrdinalIgnoreCase)
+                    ? "The uploaded CV file could not be read or was empty."
+                    : "The document extraction service could not read this CV safely.");
         }
 
         text = RepairKnownMojibake(text);
@@ -506,8 +510,12 @@ public sealed partial class ProfileCvController : ControllerBase
         {
             var generationFailure = ex as AiGenerationException;
             var retryable = generationFailure?.Retryable == true;
+            var failureMessage = generationFailure?.Message
+                ?? (ex is InvalidOperationException
+                    ? ex.Message
+                    : "CV processing failed unexpectedly. Please try again.");
             run.Status = retryable ? "queued" : "failed";
-            run.ErrorMessage = ex.Message;
+            run.ErrorMessage = failureMessage;
             run.CompletedAtUtc = retryable ? null : DateTimeOffset.UtcNow;
             await _db.SaveChangesAsync(cancellationToken);
             if (!retryable)
@@ -519,7 +527,7 @@ public sealed partial class ProfileCvController : ControllerBase
             return new CvProcessingOutcome(
                 false,
                 generationFailure?.Category ?? "cv_processing_failed",
-                ex.Message,
+                failureMessage,
                 retryable,
                 generationFailure?.Provider,
                 generationFailure?.Model,

docs/audits/verification-log.md

@@ -221,3 +221,4 @@ Output was reduced to filenames and commit counts. The token artifact appears un
 | V-187 | Focused public-CV/rate-key tests, focused authenticated-preview Jest, full backend, production frontend build and tracked-tree secret-pattern scan | Repository root / `job-tracker-ui` | Close the remaining low-risk public PDF, preview sandbox and tracked JWT hardening findings | PASS — backend focused 4/4 and full 681/681; preview Jest 15/15; production build passes. Two clients receive independent same-slug PDF partitions, both authenticated preview iframes disable scripts with a sandbox, and the values-suppressed tracked-tree scan finds no JWT/private-key pattern outside audit/operations records | No stress test, production request, history rewrite or Data Protection inspection. A coordinated history rewrite remains outside this change | JT-023/JT-025 current-tree gaps closed; expired JT-020 artifact removed from current tree |
 | V-188 | 75-user query-count fixture, 205-message/two-tenant pagination fixture, focused/full backend and frontend, optimized build | Repository root / `job-tracker-ui` | Remove JT-021's confirmed N+1 and silent 200-message ceiling without breaking old clients | PASS — admin list performs two reads independent of 75 users; page 3 returns the final 5 of 205 owned messages and excludes another tenant. Focused backend 9/9, correspondence UI 15/15, backend 683/683, frontend 58 suites/239 tests and build pass. UI page navigation and filter reset are covered; the legacy inbox endpoint remains unchanged | Synthetic SQLite/InMemory/JSDOM only; no provider mailbox, production dataset or p95 load test. Current page linked/inbound chips intentionally describe the visible page | JT-021 repository defects closed; provider/production capacity remains operational evidence |
 | V-189 | Fresh Python 3.12 virtual environments; `pytest -q`; `npm run lint`; full frontend Jest; `npm run build`; `npm audit --audit-level=moderate` | Repository root / `tools/summarizer` / `job-tracker-ui` | Restore the local Python test environment and establish a reproducible zero-warning frontend lint gate | PASS — Python 3.12.10; sidecar 26/26; ESLint zero findings; frontend 64 suites/272 tests; optimized build/TypeScript; npm audit zero vulnerabilities | Five existing SWIG deprecation warnings; Jest retains its documented force-exit/open-handle notice. ESLint 9 is intentionally pinned because the Next 16 React plugin is not ESLint 10 compatible | Local Python/frontend quality gates restored; parser dependency remediation proceeds separately under SEC-006 |
+| V-190 | Compatible parser dependency resolution; Linux CPU hash lock; clean hash install; `pip-audit`; generated parser boundary tests; focused/full backend and sidecar suites; Docker daemon probe | Repository root / `tools/summarizer` | Remove reachable upload-parser advisories and close unsafe fallback/resource-boundary paths without parsing hostile fixtures | PASS/PARTIAL — FastAPI 0.141.1/Starlette 1.6.0, Pillow 12.3.0, pypdf 6.16.2 and python-multipart 0.0.32 resolve and install from hashes; upload-facing packages audit clear; parser 32/32, focused backend 51/51 and backend 719/719. Signature/container/page/pixel/decompression/output limits pass; backend binary fallback is removed and unexpected failures are sanitized | Docker daemon unavailable, so production image smoke/container assertions did not run. `pip-audit` reports 45 Torch/Transformers model-stack advisories tracked separately under JT-017. One Starlette/httpx test-client warning and five SWIG warnings remain | SEC-006 implemented locally; SEC-007 boundary work in progress, with child-process and container isolation still required |

docs/verification/sec-006-parser-dependencies.md

@@ -0,0 +1,40 @@
+# SEC-006 parser dependency verification
+
+## Result
+
+The libraries reached by untrusted CV upload parsing were upgraded to compatible fixed releases and locked with hashes for the Linux CPU image. The dependency-only package is implemented locally but is not production-verified because the local Docker Desktop daemon is unavailable.
+
+## Fixed parser stack
+
+- Python base: `3.12.10-slim-bookworm`
+- FastAPI: `0.141.1`
+- Starlette: `1.6.0`
+- Pillow: `12.3.0`
+- pypdf: `6.16.2`
+- python-multipart: `0.0.32`
+- Reproducible Linux CPU dependency set: `tools/summarizer/requirements-linux.lock`
+
+The lock was generated for Python 3.12 on `x86_64-unknown-linux-gnu` with the PyTorch CPU index and contains hashes for every resolved package. A clean local environment installed the generated set with `--require-hashes` and imported the upgraded parser packages successfully.
+
+## Advisory review
+
+`pip-audit -r requirements.txt` reports no advisory for FastAPI, Starlette, Pillow, pypdf, PyMuPDF, python-docx, pytesseract or python-multipart. It reports 45 advisories across `torch==2.6.0` and `transformers==4.48.3`.
+
+Those remaining findings belong to the model-loading stack rather than the CV file decoder paths remediated by SEC-006. They remain open under JT-017; this record does not accept or close them.
+
+## Regression proof
+
+- Python parser suite: 32/32 passed.
+- Focused CV extraction/backend suite: 51/51 passed.
+- Full backend suite: 719/719 passed.
+- Generated boundary cases cover mismatched signatures, invalid text bytes, excessive PDF pages, excessive image pixels, excessive DOCX entries and sanitized unexpected parser failures.
+- The API no longer performs local PDF, DOCX or image parsing when the sidecar returns no text. A bounded UTF-8 text/Markdown fallback remains.
+
+## Remaining verification
+
+`docker info` fails because `dockerDesktopLinuxEngine` is not running. When a daemon is available:
+
+1. Build `tools/summarizer/Dockerfile` without changing the lock.
+2. Start the image with the intended SEC-007 runtime restrictions.
+3. Run health plus benign TXT, Markdown, PDF, DOCX and image extraction smoke tests.
+4. Record the image digest and measured memory/CPU use before any production activation.

docs/work-programmes/master-progress.md

@@ -44,7 +44,7 @@ Updated: 2026-08-30
 
 ### In progress
 
-- SEC-006 parser dependency remediation is unblocked now that package-index access has been explicitly authorized. Resolve a compatible fixed parser stack, audit it, and rerun the benign extraction corpus before starting SEC-007 isolation.
+- SEC-007 bounded document processing. Upload-facing parser dependencies are fixed and hash-locked; the sidecar now rejects signature mismatches, decompression/page/pixel/output excesses, the backend no longer retries PDF/DOCX/image parsing in-process, and unexpected parser details are sanitized. Child-process deadline/resource isolation and container controls remain.
 
 ### Remaining
 
@@ -108,17 +108,17 @@ Updated: 2026-08-30
 - Playwright: 8/10 passed on the first complete run; both failures were ambiguous selectors in the newly responsive Career selector, not product failures. Both corrected focused regressions now pass (2/2); final full rerun remains in the end-of-batch gate.
 - Manual desktop browser review: webpack development server rendered the new Career navigation and Overview correctly in dark mode; API-dependent profile status remained unavailable because the backend was not running for that isolated UI review.
 
-- **Overall programme status:** Active but externally blocked. Eight packages are locally verified and twenty-five are implemented with verification incomplete. The prioritized admin-only version indicator, every immediate repository/browser item, SEC-009, the PROD-001 read-only inventory, and the PROD-003 safe benchmark harness are complete on the release branch.
+- **Overall programme status:** Active but externally blocked. Eight packages are locally verified, twenty-six are implemented with verification incomplete, and SEC-007 is in progress. The prioritized admin-only version indicator, every immediate repository/browser item, SEC-009, the PROD-001 read-only inventory, and the PROD-003 safe benchmark harness are complete on the release branch.
-- **Current work package:** SEC-006 parser dependency remediation; package-index access is now authorized.
+- **Current work package:** SEC-007 bounded isolated document processing; SEC-006 is implemented locally and awaits Docker image smoke.
 - **Completed work packages:** None are `DONE`; all repository security/AI packages still have applicable browser, provider and/or production gates.
 - **Locally verified work:** SEC-001, SEC-002, SEC-003, SEC-005A, CORE-001, PROD-002, DEP-001 and VER-001 (`VERIFIED LOCALLY`).
-- **Implemented, verification incomplete:** SEC-004, SEC-005B, SEC-008, SEC-009, CORE-002, BG-001, OPS-001A/B/C, POL-001/002, AI-001/002/003/004, UX-001/002/003, QA-001, CAREER-001/002, MAIL-001, JOBS-001/002 and PRODUCT-001 (`IMPLEMENTED — NOT VERIFIED`). Their safe repository/browser scope is implemented; production/native-device/provider/retention gates remain where recorded.
+- **Implemented, verification incomplete:** SEC-004, SEC-005B, SEC-006, SEC-008, SEC-009, CORE-002, BG-001, OPS-001A/B/C, POL-001/002, AI-001/002/003/004, UX-001/002/003, QA-001, CAREER-001/002, MAIL-001, JOBS-001/002 and PRODUCT-001 (`IMPLEMENTED — NOT VERIFIED`). Their safe repository/browser scope is implemented; production/native-device/provider/retention gates remain where recorded.
 - **Production-verified work:** None.
 - **Blocked work:** PROD-001/003/004 and REL-001 require network/backup/model/deployment authority and unfinished dependencies. Real provider and live deletion/restore checks remain gated; DEP-001 awaits approved merge/live verification.
 - **Deferred work:** None. Conditional multi-replica coordination, model deletion, realtime operation delivery and unrelated production changes remain outside current packages.
 - **Immediate order:** all sixteen immediate repository items are complete locally, including the original UI/release queue plus SEC-009 cache/tombstone safety, worker restart clocks, universal AI accounting, email-token/Stripe lifecycle tests, exhaustive Job email selectors, the repaired migration chain, CV/public-edge hardening and measured admin/mail scaling. PROD-001 read-only evidence and the PROD-003 plan-only harness are also complete. The final audit is checking tooling/documentation before declaring only external blockers remain.
-- **Status counts:** 8 `VERIFIED LOCALLY`; 25 `IMPLEMENTED — NOT VERIFIED`; 0 `IN PROGRESS`; 1 `NOT STARTED`; 5 `BLOCKED`; 0 `DONE`; 0 `DEFERRED`.
+- **Status counts:** 8 `VERIFIED LOCALLY`; 26 `IMPLEMENTED — NOT VERIFIED`; 1 `IN PROGRESS`; 0 `NOT STARTED`; 4 `BLOCKED`; 0 `DONE`; 0 `DEFERRED`.
-- **Test status:** backend 683/683; frontend 58/58 suites and 239/239 tests; AI sidecar 23/23; Ollama benchmark harness 5/5 plus safe dry-run; optimized production build/TypeScript; EF model parity; SQLite/MariaDB migration scripts; blank/idempotent/populated SQLite migration-chain tests; disposable fresh/restarted MariaDB 11.8 application startup; Docker Compose config; safe-failure deployment preflight; and Playwright 9/9. npm audit 0 evidence remains current because the lockfile did not change. Jest's slow/open-handle behavior remains recorded.
+- **Test status:** backend 719/719; frontend 64/64 suites and 272/272 tests; ESLint zero findings; AI sidecar 32/32; Ollama benchmark harness 5/5 plus safe dry-run; optimized production build/TypeScript; EF model parity; SQLite/MariaDB migration scripts; blank/idempotent/populated SQLite migration-chain tests; disposable fresh/restarted MariaDB 11.8 application startup; Docker Compose config; safe-failure deployment preflight; and Playwright 9/9. npm audit remains at zero. The parser image build is pending because Docker Desktop's daemon is offline. Jest's slow/open-handle behavior and six Python deprecation warnings remain recorded.
 - **Deployment status:** Gitea pull-request run 609 passed the complete CI job in 4m20s. Deploy was intentionally skipped because the workflow deploys only a `push` to `main`; live remains unchanged. No merge/deployment was performed directly, no production migrations were run and the AI operation worker remains disabled by default.
 - **Production status:** State unchanged. Sanitized read-only SSH inventory was performed; no logs, prompts, private rows/content or secret values were read, and no provider/model call, model pull, service restart, file/config change, backup, restore, migration or deployment occurred. It confirmed all-interface Ollama/frontend listeners and stale database-only backups as rollout blockers.
 - **Known regressions:** None found by automated/local browser checks. Jest still needs `--forceExit` and reports its existing open-handle notice. Email-provider/send tests are fake/local only; real delivery is not claimed. Current MAIL browser evidence is 1280×720 only because the browser surface could not resize or perform native Tab traversal. Interrupted attempts are aged after 15 minutes and notified without retry; the five-minute scan is unmeasured on a large ledger. The direct clean EF-only SQLite defect and synchronous AI accounting gap are closed; migration/reconciler dual ownership remains architectural debt.

docs/work-programmes/master-work-plan.md

@@ -222,11 +222,11 @@ This queue records the highest-value work that can proceed without production cr
 - **Required tests:** dependency resolution/audit, Python suite, generated benign corpus.
 - **Required browser verification:** none for dependency-only package.
 - **Required production verification:** image digest and smoke before activation.
-- **Status:** `BLOCKED`.
+- **Status:** `IMPLEMENTED — NOT VERIFIED`.
-- **Blocker:** repository instructions prohibit internet/package resolution without explicit permission; fixed-version compatibility cannot be resolved or verified offline.
+- **Blocker:** Docker Desktop is installed but its Linux daemon is offline, so the hash-locked production image has not been built locally.
-- **Evidence:** audit lists reachable Pillow/pypdf/multipart/Starlette advisories and compatibility conflict.
+- **Evidence:** `docs/verification/sec-006-parser-dependencies.md`; upload-facing dependency audit is clear; clean hash install succeeds; parser suite 32/32; backend 719/719.
 - **Commit:** none.
-- **Remaining work:** explicit package-index permission, compatible fixed-version resolution, lock/hash refresh, audit, benign corpus parity and image smoke; do not execute malicious files.
+- **Remaining work:** build the production image from `requirements-linux.lock` and run its health/benign extraction smoke when a Docker daemon is available. Torch/Transformers advisories remain separately tracked under JT-017 and are not accepted as upload-parser findings.
 
 ### SEC-007 — Bounded isolated document processing
 
@@ -240,11 +240,11 @@ This queue records the highest-value work that can proceed without production cr
 - **Required tests:** generated boundary/corrupt fixtures, harmless sleeping child, cancellation/restart cleanup, outage/no-fallback, container assertions.
 - **Required browser verification:** synthetic CV upload status/failure; authorized private CV local-only only after safeguards.
 - **Required production verification:** measured memory/CPU limits and canary synthetic extraction.
-- **Status:** `NOT STARTED`.
+- **Status:** `IN PROGRESS`.
-- **Blocker:** follows dependency update; production sizing requires access.
+- **Blocker:** production sizing requires access; local Docker container assertions require a running daemon.
-- **Evidence:** audit parser call path and limits design.
+- **Evidence:** signature/container validation, page/pixel/decompression/output ceilings and stable parser errors have boundary tests; backend binary fallback is removed and raw unexpected failures are sanitized. Focused backend 51/51, full backend 719/719 and parser 32/32 pass.
 - **Commit:** none.
-- **Remaining work:** split behavioral and container commits if needed.
+- **Remaining work:** move decode into a deadline-bounded child process/process group, add descendant termination and cleanup tests, bound scheduling/backpressure, then apply and verify non-root/read-only/PID/CPU/memory/tmpfs container controls.
 
 ### SEC-008 — Recoverable attachment mutations
 

tools/summarizer/Dockerfile

@@ -1,4 +1,4 @@
-FROM python:3.11
+FROM python:3.12.10-slim-bookworm
 ENV PIP_NO_CACHE_DIR=1 \
     PIP_DISABLE_PIP_VERSION_CHECK=1 \
     PYTHONUNBUFFERED=1 \
@@ -8,9 +8,9 @@ WORKDIR /app
 RUN apt-get update \
     && apt-get install -y --no-install-recommends tesseract-ocr tesseract-ocr-eng \
     && rm -rf /var/lib/apt/lists/*
-COPY requirements.txt ./
+COPY requirements-linux.lock ./
 RUN python -m pip install --upgrade pip setuptools wheel \
-    && python -m pip install --extra-index-url https://download.pytorch.org/whl/cpu -r requirements.txt
+    && python -m pip install --require-hashes --extra-index-url https://download.pytorch.org/whl/cpu -r requirements-linux.lock
 COPY . .
 EXPOSE 8001
 CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8001"]

tools/summarizer/README.md

@@ -17,7 +17,7 @@ Windows:
 ```powershell
 python -m venv .venv
 .\.venv\Scripts\Activate.ps1
-pip install -r requirements.txt
+pip install -r requirements-dev.txt
 python -m uvicorn app:app --host 127.0.0.1 --port 8001 --workers 1
 ```
 
@@ -26,7 +26,7 @@ Linux / macOS:
 ```bash
 python3 -m venv .venv
 source .venv/bin/activate
-pip install -r requirements.txt
+pip install -r requirements-dev.txt
 python -m uvicorn app:app --host 127.0.0.1 --port 8001 --workers 1
 ```
 

tools/summarizer/app.py

@@ -17,6 +17,7 @@ import torch
 import pytesseract
 import threading
 import time
+import zipfile
 from urllib import request as urllib_request
 from urllib.error import URLError, HTTPError
 from contextvars import ContextVar
@@ -91,7 +92,17 @@ async def require_service_token(request: Request, call_next):
 MODEL_NAME = "sshleifer/distilbart-cnn-12-6"
 MAX_INPUT_CHARS = 20000
 MAX_CONTEXT_CHARS = 2200
-MAX_EXTRACT_FILE_BYTES = 8 * 1024 * 1024
+MAX_EXTRACT_FILE_BYTES = 5 * 1024 * 1024
+MAX_EXTRACT_CHARS = 200_000
+MAX_PDF_PAGES = 40
+MAX_DOCX_ENTRIES = 256
+MAX_DOCX_UNCOMPRESSED_BYTES = 32 * 1024 * 1024
+MAX_DOCX_ENTRY_BYTES = 8 * 1024 * 1024
+MAX_DOCX_COMPRESSION_RATIO = 100
+MAX_IMAGE_DIMENSION = 12_000
+MAX_IMAGE_PIXELS = 40_000_000
+MAX_PDF_PAGE_PIXELS = 12_000_000
+MAX_PDF_OCR_PIXELS = 120_000_000
 OCR_LANGUAGES = "eng"
 IMAGE_EXTENSIONS = {".png", ".jpg", ".jpeg", ".webp"}
 OLLAMA_BASE_URL = os.getenv("OLLAMA_BASE_URL", "http://127.0.0.1:11434").rstrip("/")
@@ -1068,36 +1079,110 @@ def _ocr_image(image: Image.Image) -> str:
     return _normalize_text(text)
 
 
+def _check_extracted_size(text: str) -> str:
+    if len(text) > MAX_EXTRACT_CHARS:
+        raise HTTPException(status_code=422, detail="The document contains too much extracted text.")
+    return text
+
+
+def _check_image_size(image: Image.Image, *, pixel_limit: int | None = None) -> None:
+    pixel_limit = MAX_IMAGE_PIXELS if pixel_limit is None else pixel_limit
+    width, height = image.size
+    if width <= 0 or height <= 0 or width > MAX_IMAGE_DIMENSION or height > MAX_IMAGE_DIMENSION:
+        raise HTTPException(status_code=422, detail="The document image dimensions are not supported.")
+    if width * height > pixel_limit:
+        raise HTTPException(status_code=422, detail="The document image contains too many pixels.")
+    if getattr(image, "n_frames", 1) != 1:
+        raise HTTPException(status_code=422, detail="Multi-frame document images are not supported.")
+
+
+def _validate_docx_container(data: bytes) -> None:
+    try:
+        with zipfile.ZipFile(io.BytesIO(data)) as archive:
+            entries = archive.infolist()
+            names = {entry.filename for entry in entries}
+            if "[Content_Types].xml" not in names or "word/document.xml" not in names:
+                raise HTTPException(status_code=400, detail="The uploaded file does not match its DOCX extension.")
+            if len(entries) > MAX_DOCX_ENTRIES:
+                raise HTTPException(status_code=422, detail="The DOCX contains too many files.")
+            total_size = 0
+            for entry in entries:
+                total_size += entry.file_size
+                if entry.file_size > MAX_DOCX_ENTRY_BYTES:
+                    raise HTTPException(status_code=422, detail="The DOCX contains an oversized file.")
+                if entry.file_size > 0 and entry.compress_size == 0:
+                    raise HTTPException(status_code=422, detail="The DOCX compression ratio is not supported.")
+                if entry.compress_size > 0 and entry.file_size / entry.compress_size > MAX_DOCX_COMPRESSION_RATIO:
+                    raise HTTPException(status_code=422, detail="The DOCX compression ratio is not supported.")
+            if total_size > MAX_DOCX_UNCOMPRESSED_BYTES:
+                raise HTTPException(status_code=422, detail="The DOCX expands beyond the supported size.")
+    except HTTPException:
+        raise
+    except (zipfile.BadZipFile, OSError) as exc:
+        raise HTTPException(status_code=400, detail="The uploaded file does not match its DOCX extension.") from exc
+
+
+def _validate_file_signature(extension: str, data: bytes) -> None:
+    matches = {
+        ".pdf": data.startswith(b"%PDF-"),
+        ".docx": data.startswith(b"PK"),
+        ".png": data.startswith(b"\x89PNG\r\n\x1a\n"),
+        ".jpg": data.startswith(b"\xff\xd8\xff"),
+        ".jpeg": data.startswith(b"\xff\xd8\xff"),
+        ".webp": len(data) >= 12 and data.startswith(b"RIFF") and data[8:12] == b"WEBP",
+    }
+    if extension in matches and not matches[extension]:
+        raise HTTPException(status_code=400, detail=f"The uploaded file does not match its {extension[1:].upper()} extension.")
+    if extension in {".txt", ".md"} and b"\x00" in data:
+        raise HTTPException(status_code=400, detail="The uploaded text file contains binary data.")
+
+
 def _extract_pdf_text(data: bytes) -> tuple[str, bool, int]:
     page_count = 0
     extracted_pages = []
     try:
         reader = PdfReader(io.BytesIO(data))
         page_count = len(reader.pages)
+        if page_count > MAX_PDF_PAGES:
+            raise HTTPException(status_code=422, detail="The PDF contains too many pages.")
         for page in reader.pages:
             try:
                 extracted_pages.append(page.extract_text(extraction_mode="layout") or "")
             except TypeError:
                 extracted_pages.append(page.extract_text() or "")
+    except HTTPException:
+        raise
     except Exception:
         extracted_pages = []
 
-    text = _normalize_text("\n".join(extracted_pages))
+    text = _check_extracted_size(_normalize_text("\n".join(extracted_pages)))
     if len(text) >= 80:
         return text, False, page_count
 
     doc = fitz.open(stream=data, filetype="pdf")
     page_count = doc.page_count
+    if page_count > MAX_PDF_PAGES:
+        doc.close()
+        raise HTTPException(status_code=422, detail="The PDF contains too many pages.")
     ocr_pages = []
+    total_pixels = 0
     for page in doc:
         pix = page.get_pixmap(matrix=fitz.Matrix(2, 2), alpha=False)
+        page_pixels = pix.width * pix.height
+        total_pixels += page_pixels
+        if page_pixels > MAX_PDF_PAGE_PIXELS or total_pixels > MAX_PDF_OCR_PIXELS:
+            doc.close()
+            raise HTTPException(status_code=422, detail="The PDF requires too much OCR processing.")
         image = Image.open(io.BytesIO(pix.tobytes("png")))
+        _check_image_size(image, pixel_limit=MAX_PDF_PAGE_PIXELS)
         ocr_pages.append(_ocr_image(image))
+        _check_extracted_size("\n".join(ocr_pages))
     doc.close()
-    return _normalize_text("\n".join(ocr_pages)), True, page_count
+    return _check_extracted_size(_normalize_text("\n".join(ocr_pages))), True, page_count
 
 
 def _extract_docx_text(data: bytes) -> str:
+    _validate_docx_container(data)
     document = Document(io.BytesIO(data))
     parts = []
     blocks = document.iter_inner_content() if hasattr(document, "iter_inner_content") else document.paragraphs
@@ -1113,11 +1198,15 @@ def _extract_docx_text(data: bytes) -> str:
             if "role" in style and parts:
                 parts.append("")
             parts.append(f"- {text}" if "bullet" in style else text)
-    return _normalize_text("\n".join(parts))
+    return _check_extracted_size(_normalize_text("\n".join(parts)))
 
 
 def _extract_plain_text(data: bytes) -> str:
-    return _normalize_text(data.decode("utf-8", errors="ignore"))
+    try:
+        decoded = data.decode("utf-8-sig")
+    except UnicodeDecodeError as exc:
+        raise HTTPException(status_code=400, detail="The uploaded text file is not valid UTF-8.") from exc
+    return _check_extracted_size(_normalize_text(decoded))
 
 
 @app.post("/extract-text")
@@ -1135,6 +1224,9 @@ async def extract_text(file: UploadFile = File(...)):
         raise HTTPException(status_code=400, detail="The uploaded file was empty.")
     if len(data) > MAX_EXTRACT_FILE_BYTES:
         raise HTTPException(status_code=400, detail="The uploaded file is too large for AI extraction.")
+    if extension not in {".txt", ".md", ".docx", ".pdf", *IMAGE_EXTENSIONS}:
+        raise HTTPException(status_code=400, detail="This file type is not supported for AI extraction.")
+    _validate_file_signature(extension, data)
 
     try:
         if extension in {".txt", ".md"}:
@@ -1149,15 +1241,14 @@ async def extract_text(file: UploadFile = File(...)):
             text, ocr_used, page_count = _extract_pdf_text(data)
         elif extension in IMAGE_EXTENSIONS:
             image = Image.open(io.BytesIO(data))
+            _check_image_size(image)
             text = _ocr_image(image)
             ocr_used = True
             page_count = 1
-        else:
-            raise HTTPException(status_code=400, detail="This file type is not supported for AI extraction.")
     except HTTPException:
         raise
     except Exception as exc:
-        raise HTTPException(status_code=500, detail=f"AI extraction failed: {exc}") from exc
+        raise HTTPException(status_code=422, detail="Document extraction failed.") from exc
 
     if not text:
         raise HTTPException(status_code=422, detail="AI extraction did not find readable text in the uploaded file.")

tools/summarizer/requirements.txt

@@ -1,12 +1,13 @@
-fastapi==0.115.12
+fastapi==0.141.1
+starlette==1.6.0
 uvicorn[standard]==0.34.0
 transformers==4.48.3
 cachetools==5.5.2
 pydantic==2.10.6
 torch==2.6.0
-pillow==11.1.0
+pillow==12.3.0
 pytesseract==0.3.13
-pypdf==5.4.0
+pypdf==6.16.2
 pymupdf==1.25.5
 python-docx==1.1.2
-python-multipart==0.0.20
+python-multipart==0.0.32

tools/summarizer/tests/test_app.py

@@ -2,9 +2,12 @@ import importlib
 import io
 import json
 import sys
+import zipfile
 from pathlib import Path
 
 from fastapi.testclient import TestClient
+from PIL import Image
+from pypdf import PdfWriter
 
 
 ROOT = Path(__file__).resolve().parents[1]
@@ -523,6 +526,86 @@ def test_extract_text_rejects_oversized_upload_before_parsing(monkeypatch):
     assert "too large" in response.json()["detail"].lower()
 
 
+def test_extract_text_rejects_extension_signature_mismatch(monkeypatch):
+    module = load_app_module(monkeypatch)
+    client = TestClient(module.app)
+
+    response = client.post("/extract-text", files={"file": ("resume.pdf", b"not a pdf", "application/pdf")})
+
+    assert response.status_code == 400
+    assert response.json()["detail"] == "The uploaded file does not match its PDF extension."
+
+
+def test_extract_text_rejects_binary_plain_text(monkeypatch):
+    module = load_app_module(monkeypatch)
+    client = TestClient(module.app)
+
+    response = client.post("/extract-text", files={"file": ("resume.txt", b"Name\x00binary", "text/plain")})
+
+    assert response.status_code == 400
+    assert response.json()["detail"] == "The uploaded text file contains binary data."
+
+
+def test_extract_text_rejects_pdf_over_page_limit(monkeypatch):
+    module = load_app_module(monkeypatch)
+    monkeypatch.setattr(module, "MAX_PDF_PAGES", 1)
+    writer = PdfWriter()
+    writer.add_blank_page(width=100, height=100)
+    writer.add_blank_page(width=100, height=100)
+    payload = io.BytesIO()
+    writer.write(payload)
+    client = TestClient(module.app)
+
+    response = client.post("/extract-text", files={"file": ("resume.pdf", payload.getvalue(), "application/pdf")})
+
+    assert response.status_code == 422
+    assert response.json()["detail"] == "The PDF contains too many pages."
+
+
+def test_extract_text_rejects_image_pixel_limit_before_ocr(monkeypatch):
+    module = load_app_module(monkeypatch)
+    monkeypatch.setattr(module, "MAX_IMAGE_PIXELS", 50)
+    monkeypatch.setattr(module, "_ocr_image", lambda image: (_ for _ in ()).throw(AssertionError("OCR must not run")))
+    payload = io.BytesIO()
+    Image.new("RGB", (10, 10), "white").save(payload, format="PNG")
+    client = TestClient(module.app)
+
+    response = client.post("/extract-text", files={"file": ("resume.png", payload.getvalue(), "image/png")})
+
+    assert response.status_code == 422
+    assert response.json()["detail"] == "The document image contains too many pixels."
+
+
+def test_docx_container_rejects_excessive_entries(monkeypatch):
+    module = load_app_module(monkeypatch)
+    monkeypatch.setattr(module, "MAX_DOCX_ENTRIES", 2)
+    payload = io.BytesIO()
+    with zipfile.ZipFile(payload, "w") as archive:
+        archive.writestr("[Content_Types].xml", "types")
+        archive.writestr("word/document.xml", "document")
+        archive.writestr("word/styles.xml", "styles")
+
+    try:
+        module._validate_docx_container(payload.getvalue())
+    except module.HTTPException as exc:
+        assert exc.status_code == 422
+        assert exc.detail == "The DOCX contains too many files."
+    else:
+        raise AssertionError("Expected the DOCX entry limit to reject the container")
+
+
+def test_extract_text_does_not_leak_parser_exception(monkeypatch):
+    module = load_app_module(monkeypatch)
+    monkeypatch.setattr(module, "_extract_pdf_text", lambda data: (_ for _ in ()).throw(RuntimeError("private path C:/secret")))
+    client = TestClient(module.app)
+
+    response = client.post("/extract-text", files={"file": ("resume.pdf", b"%PDF-invalid", "application/pdf")})
+
+    assert response.status_code == 422
+    assert response.json()["detail"] == "Document extraction failed."
+    assert "secret" not in response.text
+
+
 def test_extraction_normalization_preserves_sections_and_bullets(monkeypatch):
     module = load_app_module(monkeypatch)