creating-opencode-agents

A guide for creating OpenCode agents, which are specialized AI assistants configured for particular development tasks.

In plain words
What is it for?
It helps create project or global agents, choose whether they are primary or supporting assistants, set their instructions, and configure models, tools, and permissions.
Why use it?
It explains the file format, available settings, tool access, and permissions so an agent behaves as intended.

Skill for Claude CodeCodex

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 skills/pr-pm/prpm/creating-opencode-agents
Any agent
npx skills add pr-pm/prpm --skill creating-opencode-agents
Clone the repo
git clone --depth 1 https://github.com/pr-pm/prpm

Made for: Claude Code, Codex.

Per session 34 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,986 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00034 $0.01986
Opus 5 $0.00017 $0.00993
Sonnet 5 $0.00007 $0.00397
Haiku 4.5 $0.00003 $0.00199

Measured 2d ago against content hash b27c880e2a65, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

creating-opencode-agents 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/skills/creating-opencode-agents/SKILL.md · 345 lines

How it starts

The opening of the file, as written. The whole thing — 345 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Creating OpenCode Agents

Expert guidance for creating OpenCode AI agents with proper configuration, tools, and permissions.

When to Use

Use when:

  • User asks to create an OpenCode agent
  • Need specialized AI assistant for specific tasks
  • Building domain-focused development tools
  • Configuring agent tools and permissions

Don't use for:

  • OpenCode plugins (use creating-opencode-plugins skill)
  • OpenCode slash commands (different format)
  • Generic AI prompts (not OpenCode-specific)

Quick Reference

File Location

  • Project: .opencode/agent/<name>.md
  • Global: ~/.config/opencode/agent/<name>.md

Minimal Agent Structure

---
description: Brief explanation of the agent's purpose
mode: all
---

System prompt content here...

Required Fields

  • description (string): Brief explanation of the agent's purpose (REQUIRED)

Optional Fields

Field Type Description
mode primary | subagent | all How agent can be used (default: all)
model string Override model (e.g., anthropic/claude-sonnet-4-20250514)
temperature number Response randomness 0.0-1.0
prompt string Path to prompt file: {file:./path/to/prompt.txt}
maxSteps number Maximum iterations (unlimited if unset)
tools object Enable/disable tools with boolean values
permission object Tool access: ask, allow, or deny
disable boolean Deactivate the agent

Agent Modes

  • primary: Main assistant for direct interaction, switchable via Tab key
  • subagent: Specialized assistant invoked by primary agents or @mentions
  • all: Default; usable in both contexts

Common Patterns

Code Reviewer (Read-Only)

---
description: Reviews code for best practices and security
mode: subagent
model: anthropic/claude-sonnet-4-20250514
temperature: 0.1
tools:
  write: false
  edit: false
  bash: false
  read: true
  grep: true
  glob: true
permission:
  edit: deny
  bash: deny
---

# Code Review Agent

You are an expert code reviewer with deep knowledge of software engineering principles.

## Instructions

- Check for code smells and anti-patterns
- Verify test coverage
- Ensure documentation exists
- Review error handling and security
- Suggest improvements with examples

## Review Checklist

- [ ] Code follows project conventions
- [ ] Tests are comprehensive
- [ ] No security vulnerabilities
- [ ] Documentation is clear

Read the full file on GitHub · 345 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. 2d ago First seen · 345 lines · 34 tokens per session scan A b27c880e2a65

Subscribe to this mod's changes

creating-opencode-agents is a skill published in the GitHub repository pr-pm/prpm (120 stars, last pushed 2mo ago), licensed MIT. It adds 34 tokens to every session and 1,986 once invoked, about $0.0002 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.