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 skills add withoneai/one-agent-plugin --skill deep-seekgit clone --depth 1 https://github.com/withoneai/one-agent-pluginWrote 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.
[](https://agentmods.dev/skills/withoneai/one-agent-plugin/deep-seek)<a href="https://agentmods.dev/skills/withoneai/one-agent-plugin/deep-seek"><img src="https://agentmods.dev/badge/skills/withoneai/one-agent-plugin/deep-seek/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.
<a href="https://agentmods.dev/skills/withoneai/one-agent-plugin/deep-seek"><img src="https://agentmods.dev/badge/skills/withoneai/one-agent-plugin/deep-seek.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00097 | $0.00955 |
| Opus 5 | $0.00048 | $0.00477 |
| Sonnet 5 | $0.00019 | $0.00191 |
| Haiku 4.5 | $0.00010 | $0.00096 |
Grade A, and why
deep-seek 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 5d 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.
How it starts
The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
DeepSeek through One
DeepSeek develops advanced large language models that deliver high-performance AI capabilities for natural language understanding and generation, offering cost-effective solutions comparable to leading AI technologies.
One exposes DeepSeek through four MCP tools. The table below carries real action ids from One's knowledge base, so for a common operation you can skip search and go straight to reading the action's parameters.
How to run an action
- Find the action in the table below, or call
search_one_platform_actionswith platformdeep-seekif it is not listed. - Call
get_one_action_knowledgewith the action id. Do this every time, including for actions in this table. The table gives you the id, not the parameters. - Call
execute_one_actionwith parameters copied from that knowledge.
Never guess a parameter name, a body field, or an enum value. The knowledge has the real schema, and a guessed field is either a 400 or a silent write of the wrong thing.
Before you start
Call list_one_integrations once and confirm DeepSeek is connected. If it is missing, the user has not connected it: say so and point them at https://app.withone.ai rather than reaching for raw HTTP.
Each connection carries an access field. If it reports {"policy": "methods", "methods": ["GET"]} the agent is read-only here, so plan a read-only answer instead of attempting a write that will be refused.
Before a write
Creates, updates, deletes and sends land on a real DeepSeek account and cannot be recalled. State the action and the specific target in one line before the first write in a task, and let the user stop you. Reads need no confirmation.
Actions
ChatCompletions
| Action | Method | Path | Action id |
|---|---|---|---|
| Create Chat Completion | POST | /chat/completions |
conn_mod_def::GJ2TeNMpzqg::2h7ZYJH8S8uhqPi0XnnZ9A |
FimCompletions
| Action | Method | Path | Action id |
|---|---|---|---|
| Create FIM Completion (Beta) | POST | /beta/completions |
conn_mod_def::GJ2TeVG7lPg::9zpDzCLyQzuO5tNy6GqoBw |
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.
- 5d ago First seen · 69 lines · 0 tokens per session scan A 584a46b3db02
deep-seek is a skill published in the GitHub repository withoneai/one-agent-plugin (1 stars, last pushed 19d ago), licensed MIT. It adds 97 tokens to every session and 955 once invoked, about $0.0005 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.
Other skills, from other repositories
audio-prompting
Use when crafting TTS, music, or bed prompts for any generative audio model — director style, song structure, and post-production layering.
video-prompting
Use when crafting video or motion prompts for any generative model — dramaturgy, camera, physics-safe motion, frame anchors, and clip chaining.
generation-diversity
Use when writing any generative prompt — ritual seed, explicit structure, scenario axes, and quality gates before paid API calls.
image-prompting
Use when crafting still-image prompts for any generative model — composition, identity sheets, edits, try-on, and photoreal personas.
llm-app-patterns
Production-ready patterns for building LLM applications. Covers RAG pipelines, agent architectures, prompt IDEs, and LLMOps monitoring. Use when designing AI applications, implementing RAG, building agents, or setting up LLM observability.
prompt-optimization
Improve a prompt on the evaluations workbench through a measured loop. Score the baseline first, then duplicate the target column, form a hypothesis from failing rows, edit the copy's prompt draft, run, compare pass rate and cost, and repeat until the numbers hold. Use when the user asks to optimize or improve a…