sigmaforge

Honest Sigma-rule backtest harness. Measures detection rules against real log corpora and reports two numbers per rule — recall (does it catch attacks of its ATT&CK sub-technique?) and precision / false-positives (does it fire on benign activity?) — with honesty gates that return unmeasured instead of a fake 0 or a tautological 1.0 when the data can't support a number.

[!NOTE] Learning / portfolio project, built by directing AI coding agents. Christian Huhn (photography → SOC career change) designed, reviewed, and gated the work; the implementation was AI-pair-programmed. It is an honest measurement harness, not a polished product — see Status below for exactly what works and what doesn't yet.

What problem it solves

Every SOC ships dozens to hundreds of detection rules and rarely measures them. sigmaforge answers two questions with reproducible evidence:

• Which rules are noise generators? (high false-positives on legitimate activity)
• Which rules catch nothing? (zero recall against real attacks of their technique)

Example finding from a real run: Suspicious Windows Service Tampering produced 66 false-positives on a benign corpus — every one a Ninite / TeamViewer installer, not an attack.

How it actually works

flowchart LR
    R[SigmaHQ rules] -->|partition high/critical| C[compile to one Zircolite ruleset]
    C --> E[Zircolite engine]
    A[(attack corpus<br/>sub-technique-labeled)] --> E
    B[(benign corpus<br/>Nextron + OpTC)] --> E
    E --> S[score: recall per technique<br/>+ precision/FP label-aware]
    S --> G[honesty gates<br/>floor · positive-control · no-self-review]
    G --> O[report.md + manifest]

The pipeline is one module (sigmaforge.pipeline); the CLI and the scripts/run*_backtest.py runs call the same code — there is no weaker shipped path. Two verbs, split by honesty (you cannot compute recall/precision without labels):

# pip-runnable: rules x ARBITRARY logs -> hits only (no labels -> precision/recall = unmeasured).
# Needs only the engine, no corpora.
sigmaforge hunt --rules my_rules/ --logs /path/to/logs

# the real measurement: rules x LABELED corpora -> per-technique recall + label-aware
# precision + honesty gates + report.md + manifest.json. Needs the corpora (else it
# prints a teaching error and points you at `hunt`).
sigmaforge backtest --rules my_rules/ --config sigmaforge.yaml

--rules takes a single .yml or a directory (auto-compiled). The engine + corpora are not bundled (see Reproduce a backtest); the committed reports/run*.md are inspectable without running anything.

Status

[!IMPORTANT] The log corpora are not shipped. They are large, separately licensed, and gitignored. pip install sigmaforge installs the harness code, not the data — a full end-to-end backtest needs the corpora and a local Zircolite checkout (also not bundled). The package is useful as a library / reference; the runnable pipeline needs the local setup documented in scripts/.

Install

pip install sigmaforge

Installs the harness package and the sigmaforge CLI. The detection engine (Zircolite) and the log corpora are obtained separately (see above).

Corpora used (all verified, portfolio-safe licenses)

Reproduce a backtest from a clone

The pip package is the library/CLI; a full end-to-end backtest needs the engine and corpora, which are not bundled (engine = large third-party tool; corpora = large and separately licensed). To go from a fresh clone to a runnable backtest:

uv sync --group backtest          # script deps (pyyaml, evtx, gdown)
bash scripts/setup_engine.sh      # fetch Zircolite 3.7.6 + its runtime deps

Then obtain the corpora (see the table above) into ~/sigmaforge-v0/ and build the combined samples with the scripts/build_*_benign.py / scripts/build_*_corpus.py helpers, compile the ruleset, and run a backtest:

uv run python scripts/compile_loaded_ruleset.py   # SigmaHQ rules -> one Zircolite ruleset
uv run python scripts/run7_backtest.py            # -> reports/run7.md (+ manifest)

Run scripts from the repo root (the engine working dir defaults to the current directory; override with SIGMAFORGE_HOME if you keep Zircolite/ elsewhere). The reports/run*.md + *_manifest.json are committed so the results are inspectable without re-running.

Development

Built with the Shipwright dev framework.

uv sync --dev
uv run pytest        # 113 tests (engine-dependent tests skip without Zircolite)
uv run ruff check .

License

MIT © Christian Huhn. Corpus data retains its upstream license (see table above).