Extend FLOSS to use the rendering techniques pioneered by QUANTUMSTRAND

Organization: FLARE / Mandiant

Contributor: Lakshit Verma (@vee1e)

Mentor: Moritz Raabe (@mr-tz)

Repository: mandiant/flare-floss

Table of contents

1. Project overview
2. Project status
3. Pre-GSoC work
4. Coding-period deliverables
5. Timeline of feature work
6. Reviews I contributed
7. Future work
8. Challenges
9. Acknowledgements

Project overview

FLOSS is the FLARE team's command line tool for extracting obfuscated and plaintext strings from malware. An analyst runs it against a binary and gets the readable text the malware uses: URLs, registry paths, API names, and more. That output is useful but bare: it is a flat list of strings with no context.

QuantumStrand (QS) is an experimental branch of FLOSS that adds context. It parses the binary's structure, splits it into sections and regions, and tags each string with extra meaning: which library it likely comes from, whether it is common noise, which API family it belongs to. It also renders results in a hierarchical tree instead of a flat list.

The goal of this project was to take the best of QS and make it the default behavior of FLOSS, so every user gets layout-aware, tag-enriched output without extra flags. The project also covered the supporting infrastructure: automated string database builds, a redesigned command line, result caching, and a web viewer.

This submission is split into two parts. The pre-GSoC work below landed before the coding period started in June, mostly while I was getting familiar with the QS branch. The coding-period deliverables in the second half are the core of the GSoC project.

Project status

The project is fully complete. QS was refactored into first-class FLOSS modules, the unified pipeline is the default output path, the new command line is merged, caching ships with the tool, and the web viewer is merged. A source-level review of the merge surfaced ten issues, and I fixed the code issues. QuantumStrand was then merged into master, closing the project issue. The tracking issue stays open for the remaining product decisions, which are still tracked as three open issues: the viewer tag filter, the version bump and aliases, and the deobfuscation default.

Pre-GSoC work

These PRs landed between March and May, before the coding period. They fixed QS bugs, added format parsing, and covered community bonding tasks like CI hygiene.

Format and structure parsing

• #1221: added Mach-O parsing, including universal (fat) binaries, segments, code signatures, and entitlements.
• #1230: added PE export table parsing for DLL names, exported symbols, and forwarder names.
• #1231: added PE Rich header detection, with #1233 fixing a crash on empty headers.
• #1229: made PE resource names human-readable instead of raw numeric IDs.

Correctness fixes

• #1217: fixed false negatives for zlib error strings.
• #1225: consolidated the display of adjacent tags in the output.
• #1242: honored the skip_libs setting in decoder scoring.
• #1243: tightened import taint detection.
• #1244: removed the traceback on Ctrl+C.
• #1259: fixed a crash rendering stack strings from Rust binaries.

Exploratory and community bonding

• #1224 (closed): prototyped heuristics for junk strings in Rust binaries; we agreed a regex band-aid was not the right fix.
• #1303: added a pinact workflow that verifies GitHub Actions are pinned to commit hashes, reducing supply chain risk.
• #1304: fixed a stale pinned SHA in the build workflow.
• #1301 (closed): a first, over-detailed AGENTS.md draft that we later rewrote into the slim version below.

Coding-period deliverables

1. Native ELF parsing

• #1312: added native ELF parsing. Builds a section tree, tags relocation sections so they are not misread as strings, detects executable sections, and handles XOR-decoded ELFs. A test corpus of ELF binaries was added to the testfiles submodule.

This closed the format gap left by the pre-GSoC Mach-O and PE work, so layout rendering now works across all three major formats.

2. Automated open-source string database builds

FLOSS tags strings by matching them against databases of known library strings. Those databases used to be built by hand. This work automated the whole pipeline so they stay fresh.

• #1327: added a build script and bi-weekly CI workflow that compiles open-source libraries with vcpkg, extracts their strings, converts and deduplicates them into the FLOSS database format, and opens an automated PR with build metrics when anything changes.
• #1334: added a readable text diff of database changes, embedded in the auto-generated PR descriptions.
• #1332: fixed workflow dependencies and versioning.
• #1309: moved the string databases to Git LFS, saving roughly 24 MB per shallow clone.
• #1384: fixed the OSS DB build CI by putting the repo root on PYTHONPATH so the build script can import the floss package.
• The library list was settled at around 60 open-source libraries, chosen by auditing false-positive rates against a global prevalence database built from 150,000 real-world samples.

The pipeline has run successfully against the merged code: the bi-weekly rebuild completed cleanly (workflow run) and auto-opened the database update PR (#1385), since merged.

Building the databases surfaced several bugs in the upstream lancelot project, which the pipeline depends on. I filed reproductions and shipped fixes for all of them:

• Weak external symbols crash. Two common C++ libraries (cryptopp and jsoncpp) triggered a crash on weak external symbols. I filed the reproduction and shipped the fix, which Willi merged.
• Fragile CSV output. The jh string extractor emitted hand-rolled CSV that broke on commas, quotes, newlines, and C++ mangled names. I filed the issue and replaced the CSV with JSONL output, which Willi merged.
• Deprecated zydis build config. A stale zydis dependency broke newer CMake versions, which the flare-floss review of my database build surfaced. I filed the issue and Willi fixed it.

The database builder lives in scripts/build_oss_db.py and the automated update workflow in .github/workflows/build-oss-db.yml.

3. The QS to FLOSS integration

This is the core of the project: turning QS's separate package into the default FLOSS pipeline.

• #1337: split the monolithic QS module into first-class floss/layout and floss/tags packages and unified the command line. Roughly 4,000 lines moved and rewritten with no behavior change.
• #1347: made the unified pipeline the default. Every string type flows through the same stages into one ResultDocument, and every output mode renders from that document.

A thousand-sample parity test proved the refactor did not change behavior: 905 of 905 layout outputs matched upstream QS exactly, and 918 of 918 classic-mode outputs matched master exactly.

The pipeline is coordinated in floss/pipeline.py. The layout tree, tag engine, and renderers live in floss/layout, floss/tags, and floss/render. The results schema lives in floss/results.py.

4. The redesigned command line

QS had its own set of flags that did not match FLOSS. The spec called for one coherent interface, delivered in three parts.

• #1355 (Part 1): added symmetrical include and exclude flags for string types (--string-type / --no-string-type), renamed --functions to --analyze-functions, removed the manual load flag in favor of automatic results detection, and added runtime timing fields to the results.
• #1360 (Part 2): added render-time filters by section, structure, and tag (--section, --structure, --tag, with matching --no-* forms), --query for regex filtering that preserves the tree, --max-strings to cap output, an --interesting shortcut that drops noisy tags, a new --summary output, and sorted stable JSON with structured JSON errors on stderr.
• #1367 (Part 3): added result caching (see below).

5. Result caching

• #1367: added automatic analysis caching. The cache is keyed by a SHA-256 of the sample bytes plus the FLOSS version, stored in the platform cache directory, and written atomically behind a lock. Filters apply at render time, so cached results respect any new flags. Moritz benchmarked a sample going from about 30 seconds to about 3.5 seconds on cache hit, roughly an 8 to 10x speedup for repeat runs.

The module is floss/cache.py, with tests in tests/test_cache.py.

6. The web viewer

• #1358: reworked the QS web viewer to match the unified ResultDocument schema, added light and dark themes, fuzzy search, and virtualized rendering for very large string corpora. Deployed live for iteration during development.
• #1398 (merged): fixed broken logo and favicon on GitHub Pages by switching to relative asset paths.
• #1399 (merged): added a getting-started landing page and a download button that saves the viewer as one self-contained HTML file people can keep and open offline. The download is named after the FLOSS version and git commit, and the logo and favicon are inlined as data URIs so the file works without any external requests.

The viewer lives in the viewer/ directory and is now deployed at https://mandiant.github.io/flare-floss/ on every change. #1395 prototyped serving the viewer from the standalone binary via a --server flag; after review we decided on a simpler static HTML integration, so it was closed in favor of that approach.

7. Reproducibility and CI hygiene (coding period)

• #1308: added the slimmed-down AGENTS.md so AI coding tools have the context they need before generating code in this repo.
• #1370: deprecated all qs / QuantumStrand naming, renamed the viewer, retargeted CI to master, and cleaned up stale branches, preparing the branch for the final merge.

Timeline of feature work

The coding period ran from June through August. This is how the work above landed, in order.

June

• Early June: AGENTS.md and the Git LFS migration.
• Mid-June: native ELF parsing completed the format coverage started before the coding period.

July

• Early July: the database pipeline landed, followed by dependency fixes and database-change diffs in the auto-generated PRs.
• Late July: the first big integration step, module split.

August

• Early August: the unified pipeline became the default output path.
• Mid-August: the new command line and the filters and summary. The web viewer was reworked to match the new results schema.
• Late August: result caching, the qs naming cleanup, and the final merge, which completed the project.

Post-merge

• The OSS database rebuild succeeded with its auto-opened update PR (#1385, since merged).
• A build workflow fix unblocked the CI.
• The viewer asset path fix repaired the logo on GitHub Pages.
• An offline page made the viewer downloadable as a standalone file, with a getting-started landing page and a version-stamped, fully self-contained download.
• A simpler static HTML integration replaced the server approach, closing #1395.
• A docs update brought the usage guide up to date with the new flags, the summary output, and the cache environment variables.
• A CI tweak stopped PyInstaller builds from re-running on docs-only changes.
• Shell completion scripts (#1407, open) generate tab completion for five shells directly from the argparse parser, closing #1350 once merged.

Reviews I contributed

Part of the work was reviewing others. The notable PRs I was asked to review:

• #1319: reviewed ELF parsing robustness, including fallback for corrupt binaries. I checked the new fallback logic and requested a real corrupted test file over a mock.
• #1326: reviewed the global prevalence database update. I asked whether the large size gap between the Go and Rust sub-databases was intentional before approving.
• #1366: reviewed the summary view revamp. I caught that the new filtering let untagged generic strings crowd out tagged interesting ones, and flagged now-dead code, before it merged.
• #1405: reviewed my mentor's performance optimizations for static string extraction and rendering.
• #1273, #1294, #1328, and #1240: robustness, performance, and cleanup reviews.

Future work

What to do next, roughly in order:

• Finish the UI release. The web viewer is deployed at https://mandiant.github.io/flare-floss/ on every change, and #1399 added the downloadable offline page with a getting started view. The standalone binary will ship that static HTML page instead of hosting a server (#1371).
• Cut a beta release of FLOSS 5. Once the open work merges, a beta gives people something to try and collect feedback on.
• Resolve the tracked product decisions. Viewer tag filtering, version bump and aliases, and deobfuscation default.
• Grow the cache. A v1 tradeoff: no eviction or quota on the cache directory yet, and very large result documents load fully into memory.
• Expand database coverage. More languages and runtime sub-databases (the Go and Rust ones landed) would tighten noise filtering further.

None of these block the core functionality. The new pipeline, CLI, and cache are all merged and working.

Challenges

The largest single risk was the big refactor in #1337. Moving 4,000 lines around risks quietly changing behavior. We addressed it with evidence instead of argument: a parity harness ran old and new code over a thousand random samples, and every output matched. That became the review bar for the rest of the project.

The planning document was the other hard part. A discussion doc and an implementation spec are different things, and it took a while to realize we needed the latter. Rewriting it as a spec that could be handed to coding agents unblocked the final months of the project. It also became the design reference for the new CLI and the caching module.

Acknowledgements

Thanks to my mentor Moritz Raabe for the straight talk and the support, and for reviewing far more than his share of large diffs. Thanks to Willi Ballenthin for the sharp secondary reviews. Finally, thanks to the rest of the FLARE team for their work in creating FLOSS. This project would not have happened without them.