dark-factory

dark-factory is a skill for Claude Code, Codex from ebizsoft/agentic-claude-code. It costs 210 tokens per session (1,695 once invoked), scanned A, original, MIT.

A GitHub workflow that moves ready issues through development, automated checks, review, and merging. It offers an autonomous mode where sub-agents handle the work and an interactive mode where a human collaborates with the agent.

In plain words
What is it for?
Running ready GitHub issues from implementation through review and merge, while following branch protection and the team's project-board settings.
Why use it?
It reduces the manual coordination needed to move multiple issues through a team's normal GitHub safeguards. The human can choose whether to supervise the process or participate directly.

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/ebizsoft/agentic-claude-code/dark-factory
Any agent
npx skills add ebizsoft/agentic-claude-code --skill dark-factory
Clone the repo
git clone --depth 1 https://github.com/ebizsoft/agentic-claude-code

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 dark-factory

README.md
[![agentmods](https://agentmods.dev/badge/skills/ebizsoft/agentic-claude-code/dark-factory.svg)](https://agentmods.dev/skills/ebizsoft/agentic-claude-code/dark-factory)
Your own site
<a href="https://agentmods.dev/skills/ebizsoft/agentic-claude-code/dark-factory"><img src="https://agentmods.dev/badge/skills/ebizsoft/agentic-claude-code/dark-factory.svg" alt="Measured on agentmods" height="20"></a>
Per session 210 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,695 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 $0.00210 $0.01695
Opus 5 $0.00105 $0.00847
Sonnet 5 $0.00042 $0.00339
Haiku 4.5 $0.00021 $0.00169

Measured 4d ago against content hash 9f92e1536e8f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dark-factory 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.

The scan reads SKILL.md. This mod also ships 9 executable files (scripts/catchup.sh, scripts/lib.sh, scripts/listener.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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/dark-factory/SKILL.md · 104 lines

How it starts

The opening of the file, as written. The whole thing — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.

dark-factory — one pattern, two modes

This skill runs the Dark Factory GitHub-orchestration pattern. There is one command, two modes — pick by how the human participates:

Mode Invocation Who does the work Human's role
Autonomous (default) /dark-factory [issue] Claude (team-lead) spawns a swarm of sub-agents that develop, verify, review, and merge each Ready issue end-to-end Authorizes merges; watches the board. The factory runs itself.
Interactive /dark-factory --interactive Claude is one peer developer on a human crew — authors PRs, reviews teammates' PRs Reviews, approves, merges, collaborates over Teams. The human plays a part in the factory.

Both modes are the same conceit — issues flowing through dev → review → merge behind GitHub-native gates — at two autonomy levels. Don't blend them in one run.

Always do first (both modes)

  1. Read config/team.json — repos, your identity (.self.github), roster (.members), board (.board), base branches (.base_branches), policy, and the autonomous-mode constants (.board.project_id / .status_field_id / .status_options, .autonomous.*). scripts/lib.sh exposes most of it to the shell.
  2. Credential & scope preflight — run bash scripts/preflight.sh <autonomous|interactive> and read the report before doing any work. It proves, by making real read-only API calls (not by trusting declared scopes), that gh is authenticated and the token actually has the capabilities this run needs: repo write on every .repos repo, Projects v2 access for board moves (autonomous), webhook admin for the listener, and the opt-in teams CLI when comms are on. gh auth login's default scopes include neither project nor admin:repo_hook, so board moves and webhook registration fail silently without it. If it prints RESULT: FAIL, STOP and give the user the exact remediation it emits (usually a gh auth refresh -h github.com -s … line) — do not boot the factory on credentials that don't work. RESULT: OK (with warnings) is safe to proceed; each ⚠ has a documented fallback — note them in your summary.
  3. source scripts/lib.sh — gives you REPOS, base_branch <repo>, notify_all / notify_subset / notify_user, request_reviews.sh, webhook helpers.
  4. Read reference/budget.md — the REST-vs-GraphQL rate discipline. It governs every gh call in either mode; the factory has run the GraphQL pool to zero before. Not optional.
  5. Read reference/verification.md — the Iron Law: no completion claim ("done / green / passing / clean") without fresh verification evidence. It governs every board move, agent report, and gh pr create / merge / approve, and is enforced at the PR boundary by the verify-before-completion hook.
  6. Worktree discipline (any code change): one worktree per task at /tmp/wt-<issue>, branched off fresh origin/<base>, its own dependency install; never work in the main checkout; remove it after merge. (Full rules in the mode docs.)

Read the full file on GitHub · 104 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 · 104 lines · 210 tokens per session scan A 9f92e1536e8f

Subscribe to this mod's changes

dark-factory is a skill published in the GitHub repository ebizsoft/agentic-claude-code (2 stars, last pushed 1mo ago), licensed MIT. It adds 210 tokens to every session and 1,695 once invoked, about $0.0011 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.

Related

Other skills, from other repositories

story-readiness

Validate that a story file is implementation-ready. Checks for embedded GDD requirements, ADR references, engine notes, clear acceptance criteria, and no open design questions. Produces READY / NEEDS WORK / BLOCKED verdict with specific gaps. Use when user says 'is this story ready', 'can I start on this story', 'is…

Donchitos/Claude-Code-Game-Studios · 77 tokens

loop

Full execution protocol for MODE: LOOP — the compound-engineering loop: brainstorm → plan → build → review → improve, iterating under defense-in-depth stop conditions with generator/critic separation, durable resumable state, and mandatory compounding learning capture. Loaded on demand by the architect when the loop…

ZaxbyHub/opencode-swarm · 72 tokens

projects

List all managed projects with status, branch, open PRs, and open issue counts — portfolio-level view.

me2resh/apexyard · 24 tokens

omd:issue

불만은 채팅에 남으면 사라진다. 이 스킬은 불만을 kwakseongjae/oh-my-design의 GitHub 이슈로 옮겨 처리 가능한 큐로 만들고, 그 큐를 비우는 배치 verb까지 제공한다.

kwakseongjae/oh-my-design · 157 tokens

implementation-readiness

Verify BRD-lite, PRD, SRS/FRS, UX, and test prerequisites before implementation starts.

HoangNguyen0403/agent-skills-standard · 27 tokens

project-init

Scaffold an unconfigured directory into a configured pi project. Interactive, profile-driven: previews the planned writes, then writes AGENTS.md, .pi/settings.json and prompt files — optionally also a knowledge base, an openspec/ scaffold, and user-global /.pi/agent/settings.json. Use on a bare directory, or when the…

BlackBeltTechnology/pi-agent-dashboard · 81 tokens