Skills

In shortInstall the bundled 10xGraph skill for Codex, Claude or GitHub Copilot so your coding assistant knows the framework, or give your own agents skills.

  • 4 min read
  • 11 sections
  • Updated
  • v0.9.2
  • Markdown

Skills are folders of instructions, scripts and references that follow the Agent Skills specification. 10xGraph uses them in two places: a bundled skill that teaches your coding assistant (Codex, Claude or GitHub Copilot) how to build with 10xGraph, and a runtime feature that lets your own agents load skills on demand. This page is for developers who want an assistant to write correct 10xGraph code, and for those packaging reusable agent behavior.

Start here

To teach an assistant, run agentflow skills --agent claude (or codex, github) from your project root. The sections below list exactly which files get installed. For the command itself, see Install skills and the CLI commands reference.

To give your own agents skills, read How to give an agent skills, then the Skills reference. The tutorial Skills walks through a working example from the repository.

10xGraph uses skills in two places. Both follow the Agent Skills specification, so a skill folder is portable between them:

You want to Go to
Give your own 10xGraph agents skills they load on demand, with bundled scripts and references How to give an agent skills and the Skills reference
Teach your coding assistant (Codex, Claude, GitHub Copilot) how to build with 10xGraph The rest of this page

The 10xGraph skill for coding assistants

10xGraph skills are bundled assistant instructions for coding agents such as Codex, Claude, and GitHub Copilot. They are copied from:

Text
agentflow-api/agentflow_cli/cli/templates/skills

Use these skills when you want an assistant to understand 10xGraph package boundaries, graph patterns, CLI behavior, API routes, TypeScript client conventions, testing, and production guidance while editing an 10xGraph project.

Terminal
agentflow skills --agent codex
agentflow skills --agent claude
agentflow skills --agent github

What gets installed

The same base skill bundle is copied into the assistant-specific location.

Assistant Installed files
Codex .agents/skills/agentflow/
Claude .claude/skills/agentflow/
GitHub Copilot .github/instructions/agentflow.instructions.md and .github/skills/agentflow/

Every assistant gets the same folder, copied from:

Text
agentflow-api/agentflow_cli/cli/templates/skills/agentflow

The bundle follows the Agent Skills specification. Paths inside SKILL.md are relative to the skill folder, so one copy works in every install location. Check it, or your own skills, with agentflow skills --validate <path>.

For GitHub Copilot, 10xGraph also copies:

Text
agentflow-api/agentflow_cli/cli/templates/skills/copilot/agentflow.instructions.md

That file points Copilot at the installed skill bundle under .github/skills/agentflow/.

What the skill contains

The base bundle contains:

Text
agentflow/
+-- SKILL.md
+-- references/
    +-- architecture.md
    +-- agents-and-tools.md
    +-- state-graph.md
    +-- state-and-messages.md
    +-- checkpointing-and-threads.md
    +-- dependency-injection.md
    +-- media-and-files.md
    +-- memory-and-store.md
    +-- streaming.md
    +-- production-runtime.md
    +-- api-client.md
    +-- remote-tools.md
    +-- callbacks-and-command.md
    +-- prebuilt-agents-and-tools.md
    +-- testing-and-evaluation.md
    +-- publishers-and-runtime-protocols.md
    +-- context-id-background.md
    +-- providers-and-adapters.md
    +-- security-and-validators.md
    +-- cli-commands.md
    +-- api-configuration.md
    +-- auth-and-authorization.md
    +-- api-settings-and-middleware.md
    +-- rest-api-and-errors.md
    +-- id-and-thread-name-generators.md
    +-- client-auth-and-errors.md
    +-- client-messages-invoke-stream.md
    +-- client-threads-memory-files.md

SKILL.md is the entry point. It tells the assistant when to use the 10xGraph skill, which packages exist, where the public docs live, and which reference file to read before changing a subsystem.

The reference files cover:

Area What the assistant learns
Architecture Package layout across agentflow, agentflow-api, agentflow-client, docs, and playground
Agents and graphs Agent, ToolNode, StateGraph, prebuilt agents, state, messages, tools, and handoffs
Runtime behavior Checkpointing, dependency injection, memory, media, streaming, publishers, and protocols
API and CLI agentflow init, api, play, build, skills, 10xgraph.json, auth, settings, middleware, routes, and errors
TypeScript client Auth, invoke, stream, messages, threads, memory, files, and client-side tool execution
Quality and safety Testing, evaluation, provider adapters, validators, and prompt-injection safeguards

Install for one assistant

Run from the project root:

Terminal
agentflow skills --agent codex

Supported values are codex, claude, github, or menu numbers 1, 2, 3.

If you omit --agent in an interactive terminal, 10xGraph prompts you to choose:

Text
Which agent?
- 1. Codex
- 2. Claude
- 3. GitHub

In non-interactive environments, pass --agent or --all.

Install for every assistant

Terminal
agentflow skills --all

If an installation already exists, --all skips that assistant unless you also pass --force.

Install into another project

Terminal
agentflow skills --agent claude --path ./my-agent

The command refuses to install directly into the filesystem root or your home directory. Point --path at a project folder.

Update an existing install

Terminal
agentflow skills --agent github --force

Use --force to replace an existing installed 10xGraph skill after updating the CLI.

List supported assistants

Terminal
agentflow skills --list

Options

Option Default Description
--agent, -a interactive prompt Target assistant: codex, claude, github, or menu number 1, 2, 3
--path, -p . Project directory where skills should be installed
--force, -f false Overwrite an existing install
--all false Install skills for every supported assistant
--list, -l false List supported assistants and exit
--verbose, -v false Enable verbose logging
--quiet, -q false Suppress output except errors
Last updated for v0.9.2Edit this page on GitHubReport an issue