Dependency Vulnerability Workflow
Step 1: Gather Alerts
bash# List all open Dependabot alerts sorted by severity gh api /repos/{owner}/{repo}/dependabot/alerts --jq '.[] | select(.state == "open") | {number, severity: .security_vulnerability.severity, package: .security_vulnerability.package.name, ecosystem: .security_vulnerability.package.ecosystem, summary: .security_advisory.summary}' # Filter by severity gh api "/repos/{owner}/{repo}/dependabot/alerts?severity=critical&state=open" # Get full details for a specific alert gh api /repos/{owner}/{repo}/dependabot/alerts/{alert_number}
Step 2: Assess Impact
For each alert, determine:
- Is the vulnerable code path reachable? — Does the application actually use the vulnerable function/feature of the dependency?
- Is it a direct or transitive dependency? — Transitive vulnerabilities may be harder to fix but still pose real risk.
- What is the CVSS score and exploit availability? (Score with the CVSS convention
triaging-security-findingsdefines for effective severity.) A high CVSS with a public exploit needs immediate action. A medium CVSS with no known exploit can be scheduled. - What versions are affected and what versions fix it? — Check if updating is a minor bump or a breaking change.
Step 3: Decide on Action
| Situation | Action |
|---|---|
| Fix available, minor version bump | Update immediately |
| Fix available, major version bump | Evaluate breaking changes, schedule update |
| No fix available, code path reachable | Implement workaround or replace dependency |
| No fix available, code path not reachable | Document and monitor, set review date |
| Vulnerability in transitive dependency | Use overrides/resolutions to pin fixed version |
Transitive Dependency Risk
Direct dependencies are visible in package.json or .csproj files, but transitive dependencies (dependencies of dependencies) make up the majority of the dependency tree and are often invisible.
Why transitive dependencies matter:
- A vulnerability in a deeply nested dependency is just as exploitable as one in a direct dependency
- Transitive dependencies are less likely to be actively monitored
- Updating a transitive dependency may require updating the direct dependency that pulls it in
How to investigate:
bash# npm: Show full dependency tree npm ls --all # npm: Find which direct dependency pulls in a vulnerable transitive npm ls <vulnerable-package> # .NET: List all vulnerable packages including transitive dotnet list package --vulnerable --include-transitive # .NET: Show dependency graph dotnet list package --include-transitive
Dependency Health Evaluation
When evaluating whether to adopt or keep a dependency, assess:
| Criterion | Green Flag | Red Flag |
|---|---|---|
| Maintenance | Regular commits, responsive to issues | No commits in 12+ months, unresponded issues |
| Vulnerability History | Few CVEs, quick patches | Repeated CVEs, slow response |
| Maintainer Count | Multiple active maintainers | Single maintainer, bus factor of 1 |
| Community | High download count, active users | Very low adoption for claimed scope |
| License | Compatible with project (MIT, Apache-2.0) | Restrictive or ambiguous license |
| Security Practices | Signed releases, security policy, 2FA | No security policy, no signed releases |
Aikido Integration
Aikido is Bitwarden's container and open-source dependency scanner. It continuously scans connected repositories, surfacing open-source (SCA) and container findings in the same feed as SAST, IaC, and secrets results.
Use the aikido:issues skill (aikido_issues_list) to query findings, scoping to the relevant repo and filtering issue_types to open_source (dependency CVEs) and docker_container (container image vulnerabilities). cloud_instance is a separate Aikido issue type for cloud VM/instance findings — out of scope for dependency and container review.
Use the caller-supplied Aikido evidence if you were given any. Otherwise query aikido:issues, and if it is unavailable, stop and tell the user rather than proceeding without this data.
bash# Use the aikido:issues skill, e.g.: # issue_types: ["open_source"], repo_name: "<repo>" -> dependency CVEs # issue_types: ["docker_container"], repo_name: "<repo>" -> container image vulnerabilities
Interpreting Aikido output:
- Each finding includes: issue title, type, severity (and severity label), remediation guidance, and SLA due date
- Use
out_of_sla: trueto focus on findings already past their remediation SLA - Triage follows the same Jira-based flow as other Aikido findings — see the
triaging-security-findingsskill
Platform-Specific Guidance
NuGet (.NET)
bash# Check for vulnerable packages dotnet list package --vulnerable # Include transitive dependencies dotnet list package --vulnerable --include-transitive # Check for outdated packages dotnet list package --outdated
NuGet-specific concerns:
- .NET framework packages may have different vulnerability profiles than .NET Core
PackageReferencein.csprojis preferred overpackages.configfor better transitive resolution- Use
Directory.Packages.propsfor centralized version management in multi-project solutions
npm (Node.js)
bash# Run security audit npm audit # Auto-fix where possible npm audit fix # Force fixes (may introduce breaking changes) npm audit fix --force # Check lockfile integrity npm ci # Installs exactly from lockfile, fails if lockfile is out of date
npm-specific concerns:
package-lock.jsonmust be committed and kept in sync- Use
overridesinpackage.jsonto force transitive dependency versions:json{ "overrides": { "vulnerable-package": ">=2.0.0" } } - Beware of
postinstallscripts in dependencies — they execute arbitrary code duringnpm install
SBOM Concepts
A Software Bill of Materials (SBOM) is an inventory of all components in a software artifact. Understanding SBOMs helps reason about supply chain risk:
- What it contains: Package names, versions, licenses, relationships (direct vs. transitive)
- Why it matters: Enables rapid response when a new CVE is published — immediately identify which projects are affected
- Standard formats: SPDX, CycloneDX
- GitHub integration: GitHub generates dependency graphs automatically; Dependabot uses this for alerting
Critical Rules
- Never ignore critical/high Dependabot alerts without documented justification. Even if the vulnerable code path seems unreachable, document why.
- Prefer updating over pinning. Pinning a vulnerable version and adding a workaround accumulates tech debt. Update when a fix is available.
- Evaluate the full transitive tree. A direct dependency may be safe, but its transitive dependencies may not be.
- Review new dependencies before adoption. Check health criteria above before adding any new package. More dependencies = more attack surface.
- Lock dependencies. Always commit lockfiles (
package-lock.json,packages.lock.json). Usenpm ciin CI/CD, notnpm install.

