agentmods.dev

dcode-agent-kit CLAUDE.md

Repository guidance for an AI coding agent that builds Deep Agents and dcode command-line agents from templates. It explains the project layout, coding rules, and verified Deep Agents SDK details.

In plain words
What is it for?
Useful for scaffolding new agents, editing the related skill, choosing model settings from environment variables, and applying approval checks to agents that change data.
Why use it?
It gives the agent the project context and rules needed to create consistent, self-contained agents without embedding provider names or secret keys.

Instructions file

58 +1 2mo ago A scan 459 tokens original MIT
View the source file EliaAlberti/dcode-agent-kit
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 instructions/eliaalberti/dcode-agent-kit/claude-md
Clone the repo
git clone --depth 1 https://github.com/EliaAlberti/dcode-agent-kit
Per session 459 This file is loaded in full into every session.
When invoked 459 The same file — it is already loaded in full.
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 $0.00459 $0.00459
Opus 5 $0.00230 $0.00230
Sonnet 5 $0.00092 $0.00092
Haiku 4.5 $0.00046 $0.00046

Measured 2d ago against content hash ba57772ca40b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dcode-agent-kit CLAUDE.md 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.

CLAUDE.md · 32 lines

What it actually says

dcode-agent-kit, project context

This repo is the home of the /new-dcode-agent Claude Code skill, which scaffolds Deep Agents (LangChain) agents and dcode CLI agents. A session opened here should act as a deepagents/dcode helper.

Layout

  • skills/new-dcode-agent/SKILL.md: the skill, the whole wizard, self-contained. It embeds the agent connector and templates and writes a self-contained agent into the user's current project.
  • .claude-plugin/{plugin.json, marketplace.json}: let the skill install as a plugin via a marketplace.
  • reference/deepagents-guide.md: the Deep Agents SDK + dcode CLI knowledge base.
  • assets/dak.png: the banner.

Conventions

  • Provider-agnostic: scaffolded agents read their model and provider from env (LLM_API_KEY, optional LLM_BASE_URL, LLM_MODEL). Never bake a provider, model id, or key into generated code.
  • A scaffolded SDK agent is a self-contained folder: agent.py + a sibling model.py (a same-directory import, no sys.path hacks). A mutating agent gates its tools with interrupt_on plus a checkpointer.
  • Secrets come from env / .env (gitignored). Never commit keys.

Verified essentials (deepagents 0.6.x)

  • create_deep_agent(model=..., tools=[...], system_prompt="...", ...). The prompt param is system_prompt (not instructions). model accepts "provider:model" or a BaseChatModel instance.
  • interrupt_on={...} requires a checkpointer (e.g. InMemorySaver) or it no-ops.
  • Verify SDK facts against the installed package before asserting them; the API moves.

Editing the skill

  • Keep the skill self-contained: the connector and the agent templates stay embedded in SKILL.md, so it works whether installed standalone (~/.claude/skills/) or as a plugin. No runtime dependency on this repo.
  • Keep everything em-dash-free.
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. 2d ago First seen · 32 lines · 459 tokens per session scan A ba57772ca40b

Subscribe to this mod's changes

dcode-agent-kit CLAUDE.md is an instructions file published in the GitHub repository EliaAlberti/dcode-agent-kit (58 stars, last pushed 2mo ago), licensed MIT. It adds 459 tokens to every session, about $0.0023 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.

Building agentsagent-scaffoldingai-agentsclaude-codeclaude-code-pluginclaude-code-skilldeepagents
Related

Other instructions, from other repositories

langgraph AGENTS.md

AGENTS.md instructions for langchain-ai/langgraph, covering agents instructions, corridor security analysis, libraries and dependency map.

langchain-ai/langgraph · 477 tokens

banana-claude CLAUDE.md

Instructions for AgriciDaniel/banana-claude, covering banana claude development contract, current architecture, provider state, verified 2026-08-29, development boundaries and python support.

AgriciDaniel/banana-claude · 1,667 tokens

openharness AGENTS.md

Instructions for mifunedev/openharness, covering open harness — orchestrator, what open harness is, 1. agent work stays inside the sandbox, 2. coding-harness choice does not change the workspace and 3. remote and unattended operation are normal.

mifunedev/openharness · 2,014 tokens

soothe AGENTS.md

Instructions for mirasoth/soothe, covering soothe development guide, ⚠️ critical rules, 1. implementation guides (igs), 2. config sync and 3. ecosystem first.

mirasoth/soothe · 5,108 tokens

devin-handoff AGENTS.md

Instructions for club-cog/devin-handoff, covering devin handoff — agent guide, what this is, when to hand off, finding the script and how to use.

club-cog/devin-handoff · 542 tokens

skill-forge CLAUDE.md

Instructions for AgriciDaniel/skill-forge, covering skill forge — ultimate claude code skill creator, project overview, architecture, key principles and development rules.

AgriciDaniel/skill-forge · 1,026 tokens