Skip to content

ccpm-scheduler

Deterministic Critical Chain Project Management (CCPM) scheduling as a Python library and CLI — built for humans, AI agents, and integration into custom tools.

Give ccpm-scheduler a project network (tasks, dependencies, duration estimates, resource assignments) plus resource availability, and it produces a proper Critical Chain schedule: resource-leveled, scheduled as late as possible, protected by a project buffer and feeding buffers, complete with buffer-aware Gantt charts and interactive dependency network visualizations.

Same input always yields byte-identical output — the scheduler is fully deterministic, making it scriptable, diffable, and testable.

Why This Utility?

ccpm-scheduler was created to make Critical Chain Project Management (CCPM) lightweight, open-source, and accessible to practitioners, software developers, and AI agents.

  • Free & Frictionless Onboarding: Enterprise CCPM platforms are powerful, but evaluating CCPM can often involve vendor procurement, licensing costs, or complex server setups. ccpm-scheduler allows individuals, small teams, and researchers to experiment with and learn CCPM principles completely free, using simple text/CSV files or Python scripts.
  • Deterministic & Scriptable Engine: Offers a lightweight CLI and Python API built for local automation, continuous integration, and seamless integration with AI agent workflows and custom tools (like our-planner).
  • Pathway to Commercial Solutions: This utility serves as a starting point to prove the practical benefits of CCPM—such as aggressive task estimation, resource-leveled schedules, and aggregated buffer protection. Once organizations recognize these benefits and scale up to require enterprise features (e.g., real-time multi-user collaboration, portfolio-level fever chart tracking, and battle-tested enterprise infrastructure), they can smoothly transition to commercial CCPM software products.

Key Features

  • Deterministic & Leveling: Resource leveling ensures no resource is double-booked across concurrent tasks. Late-start scheduling maximizes project flexibility.
  • Selectable Buffer Sizing: Supports three configurable buffer calculation methods for Project Buffers and Feeding Buffers:
    • Cut & Paste (cap) (Default): Sum of safety removed from tasks in the protected chain (sum(realistic - optimal)).
    • 50% of Chain (hchain): Goldratt's classic 50% rule (ceil(0.5 * sum(optimal_duration))).
    • Root-Squared Error (rsem): Statistical combination of task variances (ceil(sqrt(sum(delta^2)))).
  • CLI & Library API: Use directly from terminal scripts, Python code, or AI agents.
  • Interactive Visualization: Generates standalone interactive dependency network views (using vis-network) and Gantt chart image plots.
  • JSON & CSV Exchange Formats: Supports clean, machine-readable inputs and outputs.

Installation

Install using pip or uv:

pip install ccpm-scheduler

Or with uv:

uv add ccpm-scheduler

Quickstart (CLI)

# Validate input files
ccpm-scheduler validate tasks.csv resources.csv calendar.csv

# Build a buffered schedule using Cut & Paste buffer sizing
ccpm-scheduler build tasks.csv resources.csv --calendar calendar.csv \
    --out-dir plan --title "Website Relaunch" --buffer-method cap

# Check schedule integrity
ccpm-scheduler check plan/schedule.csv tasks.csv resources.csv calendar.csv

# Plot Gantt chart and build interactive network graph
ccpm-scheduler plot plan/schedule.csv plan/gantt.png --resources resources.csv
ccpm-scheduler graph plan/schedule.csv plan/project-network.html \
    --tasks tasks.csv --title "Website Relaunch"

Documentation Sections

  • Walkthrough Examples: Step-by-step example comparing traditional CPM vs. CCPM schedules and buffer sizing methods (cap, hchain, rsem).
  • Core Concepts: CCPM theory, two-point estimates, late-start scheduling, and resource leveling.
  • CLI Reference: Subcommands, options, exit code contract, and --json mode.
  • Data Formats & Schemas: Specification of CSV files, unified JSON Exchange Format, and output artifacts.
  • Buffer Sizing: Deep dive into the theory, mathematics, and trade-offs of the cap, hchain, and rsem buffer calculation methods.
  • Network Layout Engine: Architectural documentation of the 6-step deterministic scheduling pipeline, ALAP baseline, resource leveling rules, and buffer insertion.

  • Python API: Library functions, dataclasses, and machine-readable validation error codes.

  • AI Agent & Tool Integration: Integration patterns for AI agents and GUI applications (like our-planner).

  • Resources & References: Recommended books, articles, YouTube lectures, and external resources.