What is a cryptographic inventory?
A cryptographic inventory is a record of every place an organisation uses cryptography (algorithms, keys, certificates and protocols), with where each one was found, what it protects and whether it is quantum-vulnerable. The EU coordinated post-quantum roadmap lists cryptographic inventories among the first steps due by the end of 2026, and an inventory is the evidence base for DORA and NIS2 cryptography policies.
CRYPTAGION builds it from three sources (source code, X.509 certificates and live TLS endpoints) and exports it as a CycloneDX 1.6 CBOM.
Inspect a pinned Python CBOM example
Two scans of 243 Python files in pyca/cryptography produced matching normalized findings and a schema-valid CycloneDX 1.6 CBOM. The evidence separates library source, tests and documentation scripts, and states the Rust coverage limits.
Explore the methodology, reviewed source evidence and downloads →
You cannot migrate, govern, or audit what you have never inventoried.
Why a cryptographic inventory is now a board topic
Most organisations cannot answer a simple question: where do RSA, ECDSA, SHA-1, MD5, TLS certificates and hardcoded keys exist across our estate? Without that map, post-quantum migration cannot be scoped or budgeted, and DORA/NIS2 cryptographic-governance expectations cannot be evidenced.
What CRYPTAGION inventories
- Code — Python, JavaScript/TypeScript, Java, Go, C/C++.
- Certificates — X.509 (PEM, DER, CRT, CER), with issuer, validity and signature algorithm.
- Live TLS endpoints — negotiated protocols and cipher suites.
Every finding carries its provenance (file, line or endpoint), so the inventory is auditable, not a black box.
Anatomy of a finding
One finding says more than an aggregate chart. This is the context CRYPTAGION records for each cryptographic asset:
- Source
src/security/KeyManager.java:143·KeyManager.wrapKey()- Scope
- Production code (not test, example or vendor)
- Detected by
- Source-code static analysis
- Use case
- Encryption (key wrapping)
- Quantum status
- Vulnerable — breakable by Shor's algorithm
- Data sensitivity
- Secret · 15-year confidentiality, from
.cryptagion.yaml - HNDL exposure
- 9 years past the 2032 CRQC planning scenario
- Risk
- Critical · 100/100, with written reasoning
- Recommended transition
- ML-KEM (FIPS 203) · migrate by 2028
Illustrative finding. The fields are the ones CRYPTAGION records for every asset; the values depend on your code and your sensitivity policy.
Findings are tagged by scope, so production cryptography is never confused with test fixtures, examples or vendored libraries, and each one keeps the detection method that produced it.
More than crypto pattern matching
Generic scanners can find cryptographic APIs. CRYPTAGION uses static analysis as one input, then turns those signals into governed cryptographic assets with provenance, context, risk, migration recommendations and standard CBOM output:
- Detection — file, line and function, or certificate and TLS endpoint
- Classification — algorithm family, key size or curve, mode, padding, use case
- Context — source scope, data sensitivity and HNDL exposure
- Risk — a 0–100 score with its reasoning
- Decision — recommended replacement, migration year and roadmap wave
- Outputs — CycloneDX CBOM, roadmap, PDF report, SARIF and SIEM events
| Capability | Generic SAST rules | SBOM tools | CRYPTAGION |
|---|---|---|---|
| Find crypto API usage in source | With custom rules | Not designed for | Yes |
| Key size, curve, mode, padding | Rule-dependent | Not designed for | Yes, where detectable |
| X.509 certificate inventory | Not designed for | Not designed for | Yes |
| Live TLS endpoint inspection | Not designed for | Not designed for | Yes |
| Data sensitivity and HNDL exposure | Not designed for | Not designed for | Yes, per-path policy |
| PQC replacement mapping | Not designed for | Not designed for | Yes |
| CycloneDX 1.6 cryptographic assets | Not designed for | Software components | Yes |
| Migration roadmap | Not designed for | Not designed for | Yes, in waves |
| Cryptographic drift between releases | Requires custom diffing | Component diffs | Yes, inventory snapshots |
From inventory to decision
The inventory is scored by quantum risk and harvest-now-decrypt-later exposure, exported as a CycloneDX 1.6 CBOM, and turned into a board report and a migration roadmap. It is the foundation for crypto-agility: knowing where cryptography is used before you have to change it. It runs inside your perimeter — on-prem or air-gapped — so your source code never leaves. See deployment & data flow.
CRYPTAGION supports compliance activities; use of the product does not by itself establish regulatory compliance.
Cryptographic inventory: frequently asked questions
What should a cryptographic inventory include?
For every cryptographic asset: the algorithm, key size or curve, mode and padding where detectable; for certificates, the issuer, validity and signature algorithm; for TLS endpoints, the negotiated protocols and cipher suites. Each entry also needs its provenance (file and line, certificate or endpoint), its scope (production, test, example or vendor), what it is used for and whether it is quantum-vulnerable. Without provenance, an inventory cannot be audited or acted on.
Will a scan find every use of cryptography in our code?
We do not claim that. Coverage depends on the language and the code style, and values passed from one method to another can still be missed. In our published comparison on Keycloak, CBOMkit found production lines our first run missed; we fixed the Java rules and published the re-run, including the 4 lines it still misses. Run any tool on code like yours before you choose.
Is every detection a vulnerability?
No. Imports, algorithm lookup tables, test vectors and capability checks all count as detections, and they mean different things. Each finding is tagged by scope, so test fixtures and vendored libraries are not confused with production cryptography. In our published pyca/cryptography scan, 1,746 of the 1,804 detections were in tests and 46 in library source.
Does our source code leave our environment?
No. Discovery runs inside your perimeter, on-prem or air-gapped: static analysis, certificate parsing, TLS handshakes, risk scoring, CBOM and report generation all run locally. Code and certificate analysis is static, so the scanned code is not executed. See deployment and data flow.
How long does a first inventory take, and what does it cost?
Pricing is public. The PQC Readiness Pilot baselines one critical domain (3–5 repositories and up to 20 TLS endpoints) for €18,000, at a reference-programme rate open to the first 5 reference customers. Discovery is a €45,000 fixed fee over 60 days, covering up to 50 repositories, 100 TLS endpoints and your certificate stores. The Platform, from €80,000 per year, keeps the inventory current.
Can the inventory be used as DORA or NIS2 evidence?
Yes, as supporting evidence. Every finding carries its provenance, and the export is a schema-validated CycloneDX 1.6 CBOM. NIS2 Article 21(2)(h) asks for policies on the use of cryptography, and the DORA risk-management RTS (Delegated Regulation (EU) 2024/1774) asks for a cryptographic controls policy, key management and a register of certificates; an inventory is how you show those are applied. Use of the product does not by itself establish compliance.
See a real inventory in 30 minutes. We run discovery live on a public repository that matches your stack and you leave with the CBOM — no access to your code, no payment until you’ve seen it work.
Request a free demo →