Ariadne
Model your whole company as one graph — then query it 22 different ways.
Most enterprise-risk tools are 22 separate products with 22 data models. Ariadne is a single graph engine — five core primitives — with 22 interchangeable lenses on top. Model the enterprise once; ask a dozen critical questions.
$ ariadne lens permission_explosion
! unintended 3-hop path to Domain Admin
bob → grp:contractors → role:deploy → DomainAdmins
$ ariadne lens outage_cost --node Billing-DB
blast radius 31 services · exposure $60k / hr
$ ariadne lens digital_twin --fail region:us-east
✔ impact modelled: 20 / 60 nodes degrade
✔ 22 lenses · one topology · 1.28s
Representative output — not a live instance.
●22 analytical lenses over one topology — vendor risk, permission explosion, outage-cost, dead-code, tech-debt scoring, org bottlenecks, digital-twin what-if, and more.
●Five graph primitives power all 22 — reachability/blast-radius, path enumeration, orphan detection, centrality/SPOF, and node scoring. Not 22 codebases; one engine.
●CLI + embedded browser UI serve the same graph and JSON API locally — no frameworks, no network clients, bounded reads and timeouts.
●Offline export parsers ingest and merge exports — Active Directory, CMDB, cloud inventory, vendor registries — into one queryable map (offline files, not live API calls).
●Adversarially hardened — self-loops, malformed JSON, empty graphs and 60k-deep chains handled iteratively without stack overflow or OOM.
0
tests, 8 crates — all green
0
lenses in 1.28s (5k nodes)
~120ms
on a 100k-deep chain
0
clippy warnings
Ruststd-only HTTP serverTF-IDF / BM25 searchserdeclapCargo workspace · 8 crates
🔒
Architecture and verified results shown. Source and binaries are private — this is a capability showcase, not a distribution.
Want risk answers
you can trust?
I can build a graph-driven risk view of your estate — or something adjacent. Tell me the questions you can't currently answer.
Start a project →