Pipelines

Pipelines define multi-step execution plans. Each step runs an adapter, and steps can depend on the output of previous steps.

Definition

Pipelines are defined in netinject.toml:

[[pipeline]]
name = "full-api-scan"
description = "Recon, scan, and fuzz"

[[pipeline.steps]]
adapter = "httpx"
label = "recon"

[[pipeline.steps]]
adapter = "nuclei"
label = "scan"
depends_on = "recon"

[[pipeline.steps]]
adapter = "ffuf"
label = "fuzz"
depends_on = "recon"

Step Fields

Field	Required	Description
adapter	Yes	Adapter name: httpx, nuclei, ffuf, sqlmap, mitmproxy
label	No	Step identifier. Used by depends_on references. Defaults to step-N.
depends_on	No	Label of a step that must complete before this one runs.
config	No	Custom config override key (reserved for future use).

Dependency Resolution

Steps are sorted topologically based on depends_on:

1. Steps with no dependencies run first (using the initial target URLs).
2. Steps with dependencies wait for their parent step to complete.
3. The parent step's output URLs become the input for the dependent step.
4. Cycles are detected and reported as errors.

Data Flow

httpx (recon) discovers 50 URLs
  ├── nuclei (scan) tests those 50 URLs for vulnerabilities
  └── ffuf (fuzz) fuzzes those 50 URLs for hidden paths

The URLs found by each step are extracted from that step's findings and passed to downstream steps.

Running a Pipeline

# Run the default pipeline
netinject run --target https://api.example.com

# Run a named pipeline
netinject run --target https://api.example.com --pipeline quick-scan

# Preview without executing
netinject run --target https://api.example.com --dry-run

Multiple Pipelines

You can define multiple pipelines in one config:

[[pipeline]]
name = "quick-scan"
description = "Fast recon + scan"

[[pipeline.steps]]
adapter = "httpx"
label = "recon"

[[pipeline.steps]]
adapter = "nuclei"
label = "scan"
depends_on = "recon"

[[pipeline]]
name = "deep-scan"
description = "Full recon, scan, fuzz, and SQLi test"

[[pipeline.steps]]
adapter = "httpx"
label = "recon"

[[pipeline.steps]]
adapter = "nuclei"
label = "scan"
depends_on = "recon"

[[pipeline.steps]]
adapter = "ffuf"
label = "fuzz"
depends_on = "recon"

[[pipeline.steps]]
adapter = "sqlmap"
label = "sqli"
depends_on = "scan"

Then select with --pipeline <name>.