execute-subagent

execute-subagent is a skill for Claude Code, Codex from fokkerone/superspecs. It costs 67 tokens per session (1,305 once invoked), scanned A, original, MIT.

A workflow for assigning coding tasks to fresh subagents and reviewing their results in waves. TDD, or test-driven development, means writing a failing test before implementing the code that makes it pass.

In plain words
What is it for?
Use it to dispatch planned tasks, give each agent an isolated context, run RED-GREEN-REFACTOR development, review completed work, and pause between execution waves.
Why use it?
It structures parallel agent work while requiring tests, specification checks, code-quality reviews, and human approval between groups of tasks.

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/fokkerone/superspecs/execute-subagent
Any agent
npx skills add fokkerone/superspecs --skill execute-subagent
Clone the repo
git clone --depth 1 https://github.com/fokkerone/superspecs

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for execute-subagent

README.md
[![agentmods](https://agentmods.dev/badge/skills/fokkerone/superspecs/execute-subagent.svg)](https://agentmods.dev/skills/fokkerone/superspecs/execute-subagent)
Your own site
<a href="https://agentmods.dev/skills/fokkerone/superspecs/execute-subagent"><img src="https://agentmods.dev/badge/skills/fokkerone/superspecs/execute-subagent.svg" alt="Measured on agentmods" height="20"></a>
Per session 67 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,305 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 $0.00067 $0.01305
Opus 5 $0.00034 $0.00652
Sonnet 5 $0.00013 $0.00261
Haiku 4.5 $0.00007 $0.00130

Measured 4d ago against content hash 7ccbfce5576b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

execute-subagent 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 4d 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.

.skills/execute-subagent/SKILL.md · 170 lines

How it starts

The opening of the file, as written. The whole thing — 170 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Skill: execute-subagent

You are orchestrating subagent-driven execution.

Each subagent gets a clean 200k-token context: the spec, one task, and the codebase. Nothing else. No prior chat history. No shared state.

TDD is not a separate step after subagent development. TDD happens inside every subagent task. Each task follows RED → GREEN → REFACTOR before it is considered done. The /superspecs:tdd skill defines the cycle; this skill embeds it into every task.

Core Principles

  • Fresh context per task. Each subagent starts clean.
  • TDD per task. Every task: RED → GREEN → REFACTOR → commit. No exceptions.
  • Two-stage review per task. Every task gets reviewed for spec compliance first, then code quality.
  • Human checkpoints between waves. No wave starts without human approval.
  • Critical findings block progress. A Critical issue in code review stops the wave.

Steps

1. Read the execution plan

Read superspec/phases/<slug>-execute/plan.md and superspec/specs/<slug>/tasks.md.

Identify:

  • The current wave (which one hasn't started yet)
  • Tasks in this wave
  • Whether they're sequential or parallel

2. Prepare subagent context package

For each task in this wave, assemble the context package. This is what each subagent receives — and ONLY this:

# Subagent Context: <Task ID>

## Spec
<full contents of spec.md>

## Your Task
<single task block from tasks.md>

## Codebase State
Branch: superspec/<slug>
Relevant files: <list files the task touches>

## TDD — Non-Negotiable
You follow RED → GREEN → REFACTOR for every piece of implementation code.

**RED**
1. Write a failing test for the behavior described in your task
2. Run it — confirm it fails for the RIGHT reason (feature missing, not a syntax error)
   - If it passes immediately: the test is wrong. Rewrite it.
   - If it fails for the wrong reason: fix the test, not the implementation.

**GREEN**
3. Write the minimum code to make the test pass
   - No YAGNI additions. No refactoring yet.
4. Run the test — confirm PASS
5. Run the full test suite — confirm no regressions

**REFACTOR**
6. Clean up: duplication, naming, complexity — keep tests green throughout
7. Run the full suite after each refactor step

**COMMIT**
8. `git commit -m "task <task-id>: <description>"`

**Code written before a failing test exists gets deleted. Start over from RED.**

## Done When
<done criteria from tasks.md for this task>

Read the full file on GitHub · 170 lines

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. 4d ago First seen · 170 lines · 67 tokens per session scan A 7ccbfce5576b

Subscribe to this mod's changes

execute-subagent is a skill published in the GitHub repository fokkerone/superspecs (4 stars, last pushed 2mo ago), licensed MIT. It adds 67 tokens to every session and 1,305 once invoked, about $0.0003 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 2026-08-31.