Pipeline Visualization

April 5, 2026 · View on GitHub

Visualize your product pipeline as an ASCII directed acyclic graph (DAG).

Overview

PA·co's 7-phase workflow is a linear pipeline with branching at Evolve (multiple products can be in Evolve simultaneously). The pipeline-viz.py tool reads your state/PIPELINE.md and renders an ASCII DAG showing where each product sits.

Usage

# Auto-detect PIPELINE.md in state/ directory
python tools/pipeline-viz.py

# Explicit path
python tools/pipeline-viz.py state/PIPELINE.md

# Compact mode (no product detail table)
python tools/pipeline-viz.py --compact

# Include completed and killed products
python tools/pipeline-viz.py --history

Example Output

PA·co Pipeline DAG
Week: SPRINT

┌────────────┐  ┌────────────┐  ┌────────────┐  ┌────────────┐  ┌────────────┐  ┌────────────┐  ┌────────────┐
│  Research  │──▶│   Refine   │──▶│Post-Refine │──▶│  CEO Gate  │──▶│  Develop   │──▶│   Deploy   │──▶│   Evolve   │
└────────────┘  └────────────┘  └────────────┘  └────────────┘  └────────────┘  └────────────┘  └────────────┘

                                                ◐ NewProduct                     ● ProductA
                                                                                 ● ProductB

Legend:
  ● Active    ◐ Blocked/Awaiting    ○ Pending

Status Indicators

Symbol	Meaning	When
●	Active	Product is progressing normally
◐	Blocked/Awaiting	Waiting for CEO decision or external input
○	Pending	Queued but not started
✓	Completed	Successfully launched (with `--history`)
✗	Killed	Terminated (with `--history`)

Phase Mapping

The tool maps various status strings from PIPELINE.md to the 7 canonical phases:

PIPELINE.md Value	Mapped Phase
`RESEARCH`	Research
`REFINE`	Refine
`POST_REFINE`	Post-Refine
`PENDING_CEO`, `CEO_GATE`	CEO Gate
`APPROVED`, `DEVELOP`, `READY_FOR_FINAL_QA`	Develop
`READY_FOR_DEPLOY`, `DEPLOYED_PENDING_CEO`	Deploy
`EVOLVE_A` through `EVOLVE_D`	Evolve

Integration with Schedules

Add pipeline visualization to your standup or weekly report schedule:

# In your standup schedule script:
python tools/pipeline-viz.py --compact

This gives agents a quick visual snapshot of where all products stand before starting their session.

How It Differs from LangGraph

LangGraph provides runtime visualization of agent execution graphs (nodes = functions, edges = transitions). PA·co's pipeline visualization shows the product lifecycle (nodes = workflow phases, products = entities moving through phases). This is a higher-level view: not "how does one agent think" but "where are all our products in the development pipeline."

Requirements

Python 3.8+
No external dependencies (stdlib only)
state/PIPELINE.md in the expected markdown table format (see state-schema.md)