Understanding Blast Radius

Learn to read Jnkn's impact analysis output.

Time: 15 minutes

What is Blast Radius?

Blast radius answers: "If I change X, what else might break?"

graph TD
    ENV[env:DATABASE_URL] --> F1[connection.py]
    ENV --> F2[users.py]
    ENV --> TF[aws_db_instance.main]
    F1 --> F3[api.py]

    style ENV fill:#ff6b6b,stroke:#c92a2a
    style F1 fill:#ffd43b,stroke:#fab005
    style F2 fill:#ffd43b,stroke:#fab005
    style TF fill:#ffd43b,stroke:#fab005
    style F3 fill:#ffe066,stroke:#fab005

Changing DATABASE_URL directly impacts 3 artifacts, and transitively impacts 1 more.

Running Blast Radius

jnkn blast env:DATABASE_URL

JSON Output (Default)

{
  "source_artifacts": ["env:DATABASE_URL"],
  "total_impacted_count": 4,
  "impacted_artifacts": [
    "file://src/db/connection.py",
    "file://src/api/users.py",
    "infra:aws_db_instance.main",
    "file://src/api/api.py"
  ],
  "breakdown": {
    "code": ["file://src/db/connection.py", "file://src/api/users.py", "file://src/api/api.py"],
    "infra": ["infra:aws_db_instance.main"],
    "env": [],
    "data": []
  },
  "max_depth_reached": 2
}

Understanding the Fields

Field	Meaning
`source_artifacts`	What you're analyzing
`total_impacted_count`	Total downstream artifacts
`impacted_artifacts`	List of affected artifacts
`breakdown`	Grouped by type
`max_depth_reached`	How far the impact travels

Reading Artifact IDs

Jnkn uses prefixed IDs to identify artifact types:

Prefix	Type	Example
`env:`	Environment variable	`env:DATABASE_URL`
`file://`	Code file	`file://src/app.py`
`infra:`	Infrastructure resource	`infra:aws_rds.main`
`data:`	Data asset (dbt model)	`data:fct_orders`
`k8s:`	Kubernetes resource	`k8s:default/deployment/api`

Limiting Depth

For large graphs, limit traversal depth:

jnkn blast env:DATABASE_URL --max-depth 1

This shows only direct dependencies, not transitive ones.

Multiple Sources

Analyze multiple artifacts at once:

jnkn blast env:DATABASE_URL env:REDIS_URL

The result shows the union of all impacted artifacts.

Output Formats

JSONMarkdownPlain

jnkn blast env:X --format json

jnkn blast env:X --format markdown

## Blast Radius: env:X

**4 artifacts impacted**

| Type | Artifact |
|------|----------|
| code | file://src/db/connection.py |
| code | file://src/api/users.py |
| infra | infra:aws_db_instance.main |

jnkn blast env:X --format plain

file://src/db/connection.py
file://src/api/users.py
infra:aws_db_instance.main

Interpreting Results

Low Impact (1-3 artifacts)

Safe to proceed. Normal for leaf nodes or well-encapsulated components.

Medium Impact (4-10 artifacts)

Review carefully. Consider notifying affected teams.

High Impact (10+ artifacts)

Proceed with caution. This change affects many components:

Schedule deployment during low-traffic periods
Have rollback plan ready
Consider feature flags

Why Is Something in the Blast Radius?

Use explain to understand why two artifacts are linked:

jnkn explain env:DATABASE_URL infra:aws_db_instance.main

Confidence: 0.87 (HIGH)

Signals:
  [+0.85] token_overlap: 3/3 tokens match [database, url, main]
  [+0.10] suffix_match: both end in common pattern

Penalties:
  [-0.08] short_token: 'url' is only 3 chars