Repository navigation
Spec Secure SDLC and Supply Chain
Part of the Technical Specification. Requirement prefix: SSC. Frameworks: NIST SP 800-218 (Secure Software Development Framework, SSDF 1.1), SLSA 1.0, OpenSSF Scorecard, OWASP ASVS 5.0. The concrete tools of the process are on Development Process; this page states what they must achieve.
| SSDF group | Practice | How it is met |
|---|---|---|
| PO, Prepare the organisation | PO.1 security requirements | This specification: SEC, NFR-SEC, NFR-PRIV, THR |
| PO.3 toolchains | Pinned CI actions, reproducible builds (SSC-005, SSC-006) | |
| PO.5 secure environments | Protected default branch, least-privilege CI tokens (SSC-010), a local stack published on loopback only (SSC-019) | |
| PS, Protect the software | PS.1 protect code | Branch protection, signed commits on release tags (SSC-011) |
| PS.2 verify integrity | Signed images, provenance (SSC-007, SSC-008) | |
| PS.3 archive releases | Every release keeps its image digest, SBOM, provenance and OpenAPI document (SSC-009) | |
| PW, Produce well-secured software | PW.1 design | Threat model (TM-001) |
| PW.4 reuse vetted components | Dependency policy (SSC-001 to SSC-004) | |
| PW.5, PW.6 secure coding and build | Coding standards ([[Spec: Development Workflow | |
| PW.7, PW.8 review and test | Required review, static analysis, fuzzing, security tests ([[Spec: Test Strategy | |
| RV, Respond to vulnerabilities | RV.1 identify | Private vulnerability reporting, scanning (SSC-013, SSC-014) |
| RV.2 assess and fix | Remediation deadlines (SSC-015) | |
| RV.3 root cause | Postmortem for security incidents ([[Spec: Operations |
| ID | Requirement | Verification |
|---|---|---|
| SSC-001 | Every dependency MUST be declared in a lockfile with exact versions and integrity hashes; builds MUST fail if the lockfile is out of date. | Lint |
| SSC-002 | A new direct dependency MUST be justified in its pull request (purpose, alternatives, maintenance state) and MUST have an OpenSSF Scorecard of at least 5.0 or an ADR accepting the gap. | Review |
| SSC-003 | Licence policy: dependencies MUST carry a licence from the allow-list (MIT, Apache-2.0, BSD-2-Clause, BSD-3-Clause, ISC, MPL-2.0, Unicode-3.0, Zlib, CC0-1.0, 0BSD). GPL, AGPL, SSPL, BUSL and unlicensed packages MUST be refused by the dependency review check; LGPL MAY be accepted by ADR for dynamically linked components only. | Lint |
| SSC-004 | Dependencies MUST be updated at least monthly by automated grouped pull requests; security updates MUST be raised within one day of disclosure. | Ops |
| ID | Requirement | Verification |
|---|---|---|
| SSC-005 | CI actions and base images MUST be pinned by digest, not by tag. | Lint |
| SSC-006 | Builds MUST be reproducible from the tagged source: same source and lockfile give the same image content (timestamps normalised). | Ops |
| SSC-007 | Every released image MUST carry an SBOM in CycloneDX 1.5 or later (JSON) and SPDX 2.3 or later, attached as attestations and published as release assets. | Ops |
| SSC-008 | Every released image MUST be signed (keyless, with a transparency-log entry) and carry build provenance meeting SLSA Build Level 3 as the target (hosted, isolated build; non-falsifiable provenance); Level 2 is the minimum for the first release. | Ops |
| SSC-009 | Deployment MUST verify the signature and provenance of the image digest before it runs; an unverified image MUST NOT be deployed. | Ops |
| SSC-010 | CI tokens MUST have least privilege per job (read-only by default, write only where publishing); no long-lived cloud or registry credential MAY be stored where short-lived federation is available. | Review |
| SSC-011 | The default branch MUST be protected (required checks, required review, no force push, linear history); release tags MUST be protected from deletion and rewrite. | Ops |
| SSC-012 | The repository MUST keep an OpenSSF Scorecard of at least 7.0, checked weekly. | Ops |
| SSC-019 | A local container stack or helper script MUST publish a port only on a loopback address by default, and the database MUST NOT publish a port at all, so a developer machine exposes nothing to its network without an explicit override. | Lint |
| ID | Requirement | Verification |
|---|---|---|
| SSC-013 | Secret scanning with push protection MUST be enabled, with a custom pattern for the key prefixes vf_live_ and vf_test_; a detected live key MUST be revoked at once. |
Ops |
| SSC-014 | Static analysis (SAST), dependency vulnerability scanning and image scanning MUST run on every pull request and on the default branch weekly. | Ops |
| SSC-015 | Remediation deadlines from disclosure or detection: critical 2 days, high 7 days, medium 30 days, low with the next minor release; an exception requires an ADR with a review date. | Ops |
| SSC-016 | A security policy (SECURITY.md) MUST describe private reporting, the response target (acknowledgement within 3 working days) and coordinated disclosure (90 days by default). |
Docs |
| SSC-017 | No secret MAY be committed, baked into an image, printed by CI or included in a test fixture; test fixtures use generated, clearly fake values. | Security |
| SSC-018 | Security-relevant changes (authentication, the licence gate, cryptography, limits) MUST receive a second review by a maintainer who did not write them. | Review |
These are the checks that meet SSC-003, SSC-005, SSC-010, SSC-012, SSC-014 and SSC-017 today. Each runs on every pull request and on every push to main; the Security workflow also runs weekly (Monday). Every tool is pinned by version, and every container image and action by digest or commit (SSC-005). A tool that can write SARIF uploads it to GitHub code scanning under its own category, so its findings are listed in the Security tab and annotated on the pull request; a pull request from a fork runs every check and uploads nothing. The repository file docs/security-checks.md holds the same list next to the workflows.
| Area | Tool | Fails the build on | SARIF in code scanning |
|---|---|---|---|
| Quality gate, coverage, duplication, hotspots | SonarCloud (CI-based analysis with the merged Go coverage profile) | the quality gate of the project (check SonarCloud Code Analysis) |
no (SonarCloud's own check) |
| SAST | CodeQL for Go, security-extended queries |
an alert at error level | yes |
| SAST in the linter | golangci-lint 2.15.0 with gosec | any finding | no (log) |
| Secrets | GitHub secret scanning with push protection; gitleaks 8.30.1 over the whole history | a detected secret (redacted in the log and the report) | gitleaks: yes |
| Dependencies | Snyk Open Source (CLI 1.1308.1, checksum verified) | high or critical | yes |
| Dependencies, reachable code | govulncheck 1.9.0 | a vulnerable function that is called | yes |
| Dependencies, new in a pull request | GitHub dependency review | a high advisory, or a licence outside the SSC-003 allow-list | no (pull request comment) |
| Dependencies, updates | Dependabot version and security updates | (opens pull requests) | alerts are in the Security tab |
| Licences | go-licenses 2.0.1 over every module | a licence outside the SSC-003 allow-list | no (report artifact) |
| Image | Trivy 0.76.0 image on the built image |
HIGH or CRITICAL with a fix | yes |
| Dockerfile | Hadolint 2.15.1 | a warning or worse | yes |
| Workflows, scripts | actionlint 1.7.7 with shellcheck | any finding | no (log) |
| OpenAPI document | Spectral 6.17.0 with the ruleset of [[Spec: API Lifecycle and Governance | Spec-API-Lifecycle-and-Governance]] | a warning or worse |
| OpenAPI changes | oasdiff 1.34.0 against the base branch and the last released document | a breaking change | no (log) |
| API fuzzing | Schemathesis against the compose stack | a failed property | no (log) |
| DAST | OWASP ZAP API scan, driven by the service's OpenAPI document, against the API container of the compose stack only, on every pull request, on main and weekly |
a rule at FAIL level; warnings are in the report artifact | yes (converted from the ZAP report, located on the OpenAPI document) |
| Posture | OpenSSF Scorecard, main and weekly |
(score; target 7.0, SSC-012) | yes |
One tool per kind of check (owner decision, 2026-11-08). A second tool that reports the same class of finding is not run: it doubles every finding and spends the run on repeated work. The test that holds the SARIF uploads to the set above fails on an upload under any other category.
| Kind of check | Kept | Removed, and why |
|---|---|---|
| SAST | CodeQL security-extended, with gosec in the linter |
Semgrep: its Go and OWASP rulesets cover classes CodeQL and gosec already report; its secrets, Dockerfile and workflow rulesets repeat gitleaks, Hadolint and actionlint |
| Dependency vulnerabilities | govulncheck (reachable code) and Snyk Open Source (a required check, the whole dependency graph) | OSV-Scanner and Trivy fs: the same advisories over the same go.mod
|
| Container image | Trivy image
|
Snyk Container: the same advisories, and a second metered test per run |
| Secrets | push protection (new pushes) and gitleaks (the whole history, which push protection does not scan) | the Trivy fs and Checkov secret scanners |
| Dockerfile | Hadolint | Checkov's Dockerfile framework |
| Workflows | actionlint with shellcheck, and the workflow conformance tests (pinned actions, least privilege, timeouts) | Checkov's GitHub Actions framework |
| DAST | OWASP ZAP and Schemathesis | none: ZAP tests security rules, Schemathesis the contract of every operation |
No DAST job reaches production: ZAP and Schemathesis are pinned to the API container of a compose stack started inside the job, with synthetic fixtures (TST-010). Each scanner was proved once to turn a pull request red, on a throwaway branch with a planted vulnerable dependency, a planted secret and a planted SQL injection pattern.
ViaFrei API: a public, keyed, read-only REST API to live German mobility data. Design phase; nothing on these pages is a measurement of a running service. Data on viafrei.de and in every API answer is attributed to its publishers under their licences.
About
Design
- Domain Model
- API Design
- Architecture Overview
- Security and Access Keys
- Rate Limits and Priority
- Performance and Caching
- Configuration
- Observability
Data Model
Technical specification
-
Technical Specification
- Interface
- Scope, Conventions and References
- Protocol
- Common Envelope
- Resource Model
- Endpoints, Places
- Endpoints, Roads
- Endpoints, Energy and Parking
- Endpoints, Public Transport and Accessibility
- Endpoints, Weather Risk, Driving Rules and Composed Features
- Service Endpoints
- Errors
- Authentication and Authorisation
- Rate Limiting, Quotas and Priority
- Data Licensing and Attribution
- Compatibility and Versioning
- OpenAPI Skeleton
- Quality
- Delivery and operations
- Interface
Testing
- Testing Strategy
- Testing and Quality (design summary)
Engineering
Technology choice
- Technology Evaluation
- Architecture: Go
- Architecture: Rust
- Architecture: TypeScript
- Stack Comparison
- Benchmark Plan
Reference