auto-dev

auto-dev is a command for Claude Code from Zts0hg/codexspec. It costs 11 tokens per session (1,864 once invoked), scanned A, original, MIT.

A command that autonomously develops each pending requirement in a shared product blueprint using a requirements-first software-development process.

In plain words
What is it for?
It is for taking confirmed blueprint requirements in order, creating their specifications, implementing them, and completing the development workflow.
Why use it?
It removes the need to manually advance every confirmed requirement through planning, implementation, and finalization.

Command for Claude Code

Written for Claude Code: argument-hint in frontmatter.

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 commands/zts0hg/codexspec/auto-dev
Clone the repo
git clone --depth 1 https://github.com/Zts0hg/codexspec

Made for: Claude Code.

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 auto-dev

README.md
[![agentmods](https://agentmods.dev/badge/commands/zts0hg/codexspec/auto-dev.svg)](https://agentmods.dev/commands/zts0hg/codexspec/auto-dev)
Your own site
<a href="https://agentmods.dev/commands/zts0hg/codexspec/auto-dev"><img src="https://agentmods.dev/badge/commands/zts0hg/codexspec/auto-dev.svg" alt="Measured on agentmods" height="20"></a>
Per session 11 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,864 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 5.1 $0.00011 $0.01864
Opus 5 $0.00005 $0.00932
Sonnet 5 $0.00002 $0.00373
Haiku 4.5 $0.00001 $0.00186

Measured 2d ago against content hash 8c5b8c844e91, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

auto-dev 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/commands/codexspec/auto-dev.md · 149 lines

How it starts

The opening of the file, as written. The whole thing — 149 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Continuous Blueprint Development

Language Preference

Read .codexspec/config.yml. Converse in language.interaction and author SDD artifacts in language.document, each falling back to language.output, then English.

Goal

Continuously run the complete Requirements-First SDD flow for the shared blueprint in document order. Requirements in the blueprint are already confirmed. Do not ask the user to reconfirm them, perform new requirements discovery, or depend on workflow.auto_next.

Run Ownership and Finalization

  1. Run codexspec _auto-dev-helper acquire with {} on stdin. If another live run owns the repository, report it and exit immediately without waiting.
  2. Keep the returned opaque token. Run renew before and after each stage and tool operation. If an operation can run longer than the returned heartbeat_interval_seconds, keep calling renew at that interval until it finishes. Run assert-owner immediately before every repository mutation. Never reuse a token after an ownership failure.
  3. Inspect the returned merge_recovery. When it reports needs_resolution, this run has fenced an interrupted synchronization merge to its new token; edit every conflicted file and call prepare-sync-verification with the token and the exact complete resolved_paths list before running checks. When it reports needs_verification, use the existing merge result as the verification candidate. Run the project's required baseline checks and call continue-sync, or call abort-sync if conflict preparation or a passing baseline cannot be completed. Then repeat sync-default until it returns clean, and run the baseline checks once more before selecting any requirement.
  4. On every normal success or controlled stop, call release in finalization. An unexpected termination is recovered by stale-owner reclamation on the next invocation.
  5. blueprint remains allowed while this run is active; do not hold a blueprint or Git lock while interpreting documents or implementing code.

Read the full file on GitHub · 149 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. 2d ago First seen · 149 lines · 11 tokens per session scan A 8c5b8c844e91

Subscribe to this mod's changes

auto-dev is a command published in the GitHub repository Zts0hg/codexspec (5 stars, last pushed 4d ago), licensed MIT. It adds 11 tokens to every session and 1,864 once invoked, about $0.0001 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 2026-09-03.