A procedure for bounded software-development tasks such as implementation, bug fixes, refactoring, and focused testing. It also defines how independent tasks can be run in separate Git worktrees, which are isolated working copies of a repository.

Implementing changes, fixing bugs, refactoring code, running targeted tests, delivering development work, and coordinating explicitly parallel tasks.

It keeps development work limited to a defined task and helps prevent parallel changes from interfering with one another.

dev-flow — Skill for Claude Code by Innocent-children

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.

Any agent

npx skills add Innocent-children/dev-flow --skill dev-flow

Clone the repo

git clone --depth 1 https://github.com/Innocent-children/dev-flow

Made for: Claude Code.

Or install dev-flow-claude, the plugin that ships this one along with the rest of its 1 skill, 1 hook, 1 MCP server.

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/innocent-children/dev-flow/dev-flow/github.svg)](https://agentmods.dev/skills/innocent-children/dev-flow/dev-flow)

Your own site

<a href="https://agentmods.dev/skills/innocent-children/dev-flow/dev-flow"><img src="https://agentmods.dev/badge/skills/innocent-children/dev-flow/dev-flow/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/innocent-children/dev-flow/dev-flow"><img src="https://agentmods.dev/badge/skills/innocent-children/dev-flow/dev-flow.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 23 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 561 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector pass 7 Sept 2026

How audits are shown

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.

Model	Per session	Once invoked
Fable 5.1	$0.00023	$0.00561
Opus 5.5	$0.00009	$0.00224
Sonnet 5.5	$0.00005	$0.00112
Haiku 4.5	$0.00002	$0.00056

Measured 1mo ago against content hash 3fefe44d1d14, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

dev-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 1mo 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.

packages/claude/plugin/skills/dev-flow/SKILL.md · 39 lines

What it actually says

Dev Flow for Claude Code

Use /dev-flow-claude:dev-flow for assessment. Claude performs repository work and authorized Host operations; Core owns Task state, Actions, transitions, verification, blockers and recovery. Invocation does not authorize Git changes or approve an unseen plan. Reuse valid user answers.

Request or state	Read and perform
New development request	Admission: assess read-only, obtain missing choices, then prepare every confirmed repository; default to a new local branch.
Saved launch	Read its status/receipt and actual workspace identity through admission before creation/resume; a requested Claude session does not prove Core creation.
Existing Task	Return to the original workspace and connect/resume, without preparing another workspace or creating another Task.
Current Action	Action execution and only the current node guide.
Rejection, lost result or blocker	Response handling and its linked recovery procedure.
Installation, relocation, cancellation, abandonment or cleanup	Host lifecycle.

After provisioning or on resume, perform server_info first and require Claude support. Use actually visible tools through transport. Retain full responses, inspect ok, then recovery/blocker/outcome and the complete Action.

Follow the Action procedure for requirements/design discussion, saved-plan approval, method steps and submission. Use artifacts before writes and every submission, and verification for checks. Required user decisions stay with the user. Write/Edit/NotebookEdit require the trusted PreToolUse Hook; a denial or disabled gate cannot be bypassed with another tool. Bash/external writes are observed later. Claude permissions still apply.

Follow current user and CLAUDE.md/AGENTS.md instructions. Confirm every repository (at most eight) and its permissions before preparation; never create a partial Task. Indexes are optional and are not installed automatically. Read handoff before another session. DONE/CANCELLED preserves local work and grants no commit, push, publication or deletion authority; dedicated worktree and branch deletion require their separate authorization.

Implementation: packages/claude/lib/workspace.mjs; packages/claude/bin/dev-flow-claude.mjs; packages/claude/plugin/hooks/pre-tool-use.mjs.

Files

What ships with it

60 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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.

1mo ago Changed · +10 lines 3fefe44d1d14
1mo ago Changed · -108 lines · -53 tokens per session 96a6438cd451
1mo ago Changed · -709 lines · +10 tokens per session 5d0a89426593
1mo ago Changed · +163 lines 4ac18fbb23f4
1mo ago Changed · +8 lines 45b4a0df8d30
1mo ago Changed · +198 lines · -18 tokens per session 80f7130c5371
1mo ago First seen · 477 lines · 84 tokens per session scan A d7368fd07e5e

Subscribe to this mod's changes

dev-flow is a skill published in the GitHub repository Innocent-children/dev-flow (13 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 23 tokens to every session and 561 once invoked, about $0.0001 per session on Opus 5.5. 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.

Git and workflow codex codex-plugin codex-skills deepseek deepseek-harness-plugin development

dev-flow

Getting it into your agent

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

What it actually says

Dev Flow for Claude Code

What ships with it

What this file has done since we first saw it

Other skills, from other repositories

git-workflow-and-versioning

git-commit

make-repo-contribution

finishing-a-development-branch

final-release-review

pr-draft-summary