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.
npx agentmods add skills/0x0w1/jig/develop-task-flownpx skills add 0x0w1/jig --skill develop-task-flowgit clone --depth 1 https://github.com/0x0w1/jigWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00050 | $0.01754 |
| Opus 5 | $0.00025 | $0.00877 |
| Sonnet 5 | $0.00010 | $0.00351 |
| Haiku 4.5 | $0.00005 | $0.00175 |
Grade A, and why
develop-task-flow 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.
This is a copy
100% identical to jig-develop-task-flow — 2 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 130 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Develop Task Flow
Use this repository skill for normal development work requested by the user.
Branch Model
- Start from
origin/develop. - Create one task branch:
feature/<slug>for user-visible features or enhancements.fix/<slug>for bug, regression, or security fixes.chore/<slug>for tooling, dependencies, refactors, docs, or automation setup.
- Finish the task by squash-merging the branch into
developlocally and pushingdevelop. There are no pull requests. - Ordinary code, config, documentation, generated
dist, workflow, and installer changes must follow this flow before any release can include them. - A release promotes
developtomainwith a fast-forward push; it is not a task. Usegithub-releasefor releases.
Commit Message Rules
The squash commit on develop is the release-note source. Every squash commit must follow this format:
- Subject:
<type>: <concise user-facing summary>with<type>one offeat,fix,chore,docs,refactor,test,ci. - Release note sections derive from the prefix:
feat:renders under🚀 Enhancements,fix:under🐛 Fixes,chore:under🧰 Chores, and any other prefix becomes its own section named after it (docs:→📚 Documentation). Pick the prefix that describes the change, not the section you want. - Body: bullet items describing the change from the user's perspective, release-note ready. Write them in the language the repository already uses for its commit bodies, defaulting to English; match the existing history rather than switching it.
- In body bullets, wrap useful technical terms in backticks, such as file paths, config keys, branch names, workflow names, command names, and env vars.
Release Grade
The squash commit records the grade this task earns. Decide it here rather than at release time: at merge the diff, the tests, and the reasoning are still in hand, while a release has to reconstruct all of it from commit text.
Grade against the repository's own rubric, not a fixed prefix mapping — a prefix names the kind of change, not what it costs. Resolve the rubric path in this order:
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.
- 2d ago First seen · 130 lines · 50 tokens per session scan A 249b93771002
develop-task-flow is a skill published in the GitHub repository 0x0w1/jig (5 stars, last pushed 2d ago), licensed MIT. It adds 50 tokens to every session and 1,754 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to jig-develop-task-flow, differing in 2 lines, and is treated as a copy.
Other skills, from other repositories
git-ship
自动化 Git 工作流一键发布助手，执行完整的「ship」流程： 基于最新 main 切新分支 → commit → 推送 → 创建 PR → squash merge → 回 main。 调用 ship 本身就是对完整流程的授权；自动推断分支名、commit message 和 PR 内容， 并主动解决 Git 冲突、测试、lint、类型检查、构建和 CI 失败，持续修复和重试， 直到改动成功合并。不要因可自行修复的失败暂停，只在认证、权限、仓库保护等 无法由当前环境解决的外部阻塞出现时请求用户介入。 仅在用户明确表达「ship」意图时触发，例如： 直接说「ship」「/ship」「git ship」 明确要求完整走完 git…
wu5-dev-flow
使用可审计的 SDD、严格 RED-GREEN-REFACTOR TDD 与安全 Git 门禁初始化、开发、修复、重构和交付 Python 项目。用于任何会修改项目源码、测试、规格、依赖或 Git 历史的任务，也用于继续跨 Session 的现有 spec/ 变更、审查代码、验证完成状态、创建提交或准备 GitHub PR。.
ascii-diagram-creator-skill
Create ASCII diagrams from workflow definitions and save them as image files (PNG, SVG, etc.).
watchfire
Operate Watchfire — the local orchestrator that runs coding agents (Claude Code, Codex, Gemini, opencode, Copilot) on tasks in sandboxed git worktrees and merges the results. Use when asked to run/queue/inspect Watchfire tasks or agents, drive the factory loop (create → run → wait → review), check why an agent is…
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.