deepseek-flow

deepseek-flow is a skill for Claude Code, Codex from kanghelyu/dsh-deepseek-flow. It costs 27 tokens per session (1,880 once invoked), scanned A, original, MIT.

A workflow tool for DeepSeek Flow uses Markdown files to build, inspect, and maintain visual multi-step processes with yes-or-no decision gates.

In plain words
What is it for?
Use it to create, import, view, or update multi-step workflows and their branching logic.
Why use it?
It keeps workflow steps and decisions organized in files while preserving the connections between them.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it to create, import, view, or update multi-step workflows and their branching logic.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/kanghelyu/dsh-deepseek-flow/deepseek-flow
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 kanghelyu/dsh-deepseek-flow --skill deepseek-flow
Clone the repo
git clone --depth 1 https://github.com/kanghelyu/dsh-deepseek-flow

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 deepseek-flow

README.md
[![agentmods](https://agentmods.dev/badge/skills/kanghelyu/dsh-deepseek-flow/deepseek-flow/github.svg)](https://agentmods.dev/skills/kanghelyu/dsh-deepseek-flow/deepseek-flow)
Your own site
<a href="https://agentmods.dev/skills/kanghelyu/dsh-deepseek-flow/deepseek-flow"><img src="https://agentmods.dev/badge/skills/kanghelyu/dsh-deepseek-flow/deepseek-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.

agentmods 80×15 button for deepseek-flow

Your own site · 80×15
<a href="https://agentmods.dev/skills/kanghelyu/dsh-deepseek-flow/deepseek-flow"><img src="https://agentmods.dev/badge/skills/kanghelyu/dsh-deepseek-flow/deepseek-flow.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 27 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,880 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 warn 7 Sept 2026
SkillSpector: 1 finding, up to medium

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • medium Excessive Agency · line 10
    Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
    Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00027 $0.01880
Opus 5 $0.00014 $0.00940
Sonnet 5 $0.00005 $0.00376
Haiku 4.5 $0.00003 $0.00188

Measured 10d ago against content hash 79759a7460e6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

deepseek-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 10d 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/deepseek-flow/SKILL.md · 102 lines

How it starts

The opening of the file, as written. The whole thing — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.

DeepSeek Flow

Use this skill whenever the user wants to build, import, visualize, or modify a multi-step workflow. Once the request is clear, call the tool directly; do not ask the user to confirm details again.

Creating and updating

  • Prefer flow_create for new workflows. It creates WORKFLOW.md, one STEP.md workspace per step, and the canvas nodes and edges.
  • To update an existing workflow, first call flow_read to get the current revision, then call flow_put with that expected_revision.
  • A successful flow_create or flow_put already persists the topology. Do not ask the user to click "Apply changes" in Studio, and do not send the same topology back to the main Session for review. Studio syncs by revision automatically.
  • If you edit an existing workflow's WORKFLOW.md, STEP.md, or definition files directly (without using flow_put) and the edit changes nodes, gates, or arrows, call flow_finalize_canvas after the file changes, passing the workflow id (and preferably the expected_revision from flow_read). It triggers an invisible deterministic finalize in Studio and saves directly without another main-Session review.
  • Even if you forget flow_finalize_canvas, Studio falls back to detecting that no canvas edit event occurred and presses the same hidden finalize path automatically. Canvas drafts created by the user in Studio still require "Apply changes".
  • Ordinary execution edges must be acyclic. For bounded retries, add an explicit feedback edge: {"source":"review","target":"implement","feedback":{"maxIterations":3,"exitCondition":"quality check passed"}}. Feedback edges must have an integer limit between 1 and 1000 and a non-empty exitCondition. They do not participate in single-pass Boolean gate evaluation and do not auto-run Agent steps; the current Session must control each retry according to WORKFLOW.md.

Naming language

  • Use concise workflow names in the user's language. If the user did not specify a name, flow_create generates a default name in the language you pass.
  • Pass language: "en" or language: "zh" to flow_create so default node labels (Input, Output, Step N) and the default four steps match the user's language. If omitted, the default is Chinese for backward compatibility; you should set it to the language the user is currently using.

Read the full file on GitHub · 102 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. 10d ago First seen · 102 lines · 27 tokens per session scan A 79759a7460e6

Subscribe to this mod's changes

deepseek-flow is a skill published in the GitHub repository kanghelyu/dsh-deepseek-flow (69 stars, last pushed 21d ago), licensed MIT. It adds 27 tokens to every session and 1,880 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 2026-08-30.