graflow-workflow

A structured way to build Python workflows with Graflow, a framework for connecting tasks into executable graphs. It uses three phases: planning, implementation, and review.

In plain words
What is it for?
Use it to design task graphs, automation pipelines, parallel processing, or AI workflows in Python. It produces a design document and then guides implementation and review.
Why use it?
It turns a complex pipeline into explicit tasks, dependencies, data flows, and error-handling decisions before code is written. This helps clarify which work is sequential, parallel, conditional, or connected to an AI model.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/graflowai/graflow/graflow-workflow
Any agent
npx skills add GraflowAI/graflow --skill graflow-workflow
Clone the repo
git clone --depth 1 https://github.com/GraflowAI/graflow

Made for: Claude Code, Codex.

Per session 73 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,422 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00073 $0.01422
Opus 5 $0.00036 $0.00711
Sonnet 5 $0.00015 $0.00284
Haiku 4.5 $0.00007 $0.00142

Measured 2d ago against content hash 405e8148f6a5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

graflow-workflow scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/graflow-workflow/SKILL.md · 226 lines

How it starts

The opening of the file, as written. The whole thing — 226 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Graflow Workflow Builder

Build executable task graphs in Python using a structured 3-phase approach.

Workflow: Plan -> Implement -> Review

Phase 1: Plan

Goal: Clarify requirements, create a design document, and get user approval through iterative feedback.

Steps:

Step 1: Requirements Gathering

Ask the user clarifying questions about the workflow:

  • What is the workflow's purpose?
  • What are the input sources and output destinations?
  • Which tasks need to run sequentially vs. in parallel?
  • Is LLM integration needed?
  • Are there any dynamic/conditional branching requirements?
Step 2: Create Design Document

Create a design document (workflow_design.md) with:

  • Workflow overview and purpose
  • Task definitions (name, responsibility, inputs/outputs)
  • Task graph structure (ASCII diagram)
  • Channel data flow
  • Error handling strategy

Design Document Template:

# Workflow Design: {workflow_name}

## Overview
{Brief description of what this workflow accomplishes}

## Tasks

| Task ID | Responsibility | Inputs | Outputs |
|---------|---------------|--------|---------|
| task_a  | ...           | ...    | ...     |

## Task Graph

source >> (transform_a | transform_b) >> sink


## Channel Data Flow
- `config`: Set by setup, used by all tasks
- `results`: Accumulated by each task

## Error Handling
- {Strategy: fail-fast, best-effort, retry, etc.}
Step 3: Present Design to User

Present the design document to the user with a clear summary:

  • Show the task graph diagram
  • Highlight key design decisions
  • Ask explicitly: "Does this design meet your requirements? Please provide feedback if any changes are needed."
Step 4: Iterate on Feedback

If the user provides feedback:

  1. Update workflow_design.md with the requested changes
  2. Summarize the changes made
  3. Re-present the updated design
  4. Repeat until the user is satisfied
Step 5: Confirm Design Approval

Before proceeding to implementation:

  • Ask the user to confirm: "Is this design approved? If yes, I'll proceed to implementation."
  • Only move to Phase 2 after explicit approval

Read the full file on GitHub · 226 lines

Files

What ships with it

3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 226 lines · 73 tokens per session scan A 405e8148f6a5

Subscribe to this mod's changes

graflow-workflow is a skill published in the GitHub repository GraflowAI/graflow (41 stars, last pushed 19d ago), licensed Apache-2.0. It adds 73 tokens to every session and 1,422 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.