---
title: "Parallel coding agents with specs"
canonical: "https://agenticfeats.com/c/parallel-coding-agents-with-specs-claude-code/"
agents: ["Claude Code"]
evidence_grade: "A"
outcome: "completed"
job_type: "Coding"
roles: ["Developer"]
source_url: "https://schipper.ai/posts/parallel-coding-agents"
evidence_url: "https://gist.github.com/manuelschipper/149ebf6b2d150ccaccc84ee9a9df560f"
source_date: "2026-02-26"
published: "2026-10-13"
updated: "2026-10-13"
---

> From Agentic Feats: sourced examples of AI agents at work. Map of the site for agents: https://agenticfeats.com/llms.txt

# Parallel coding agents with specs

Claude Code agents run in separate tmux windows as planner, worker and PM, driven by numbered Markdown feature specs and six slash commands.

Result: The author ran 4 to 8 agents in parallel and wrote 300+ feature specs in one project, then ported the system to new repos.

## How it was set up

Claude Code in tmux windows, Markdown Feature Design files with an index, bash aliases, six slash commands, and Cursor for editing specs. An /fd-init command bootstraps any repo.

## Worth knowing

Beyond 8 agents the author could not keep up and their decision quality dropped.

## Prompt

```text
Read the feature design [FD-number], check the %% notes I added inline, and revise the spec. Then list the files to change and a verification plan.
```

## Evidence note

Blog post with a public gist of the slash commands; author states 300+ specs

## Sources

- Original: https://schipper.ai/posts/parallel-coding-agents
- Evidence: https://gist.github.com/manuelschipper/149ebf6b2d150ccaccc84ee9a9df560f

## The receipt

- Agent: Claude Code
- Evidence: Shown (grade A, artifact linked)
- Outcome: Completed
- Job type: Coding
- Roles: Developer
- Source: Blog
- Published: 2026-02-26
- Added: 2026-10-13