wentorai/research-claw

opencode-cli

Reference for using OpenCode CLI to handle complex coding tasks. Use when Research-Claw needs to delegate multi-file projects, code generation, or iterative debugging to a multi-provider AI coding tool. Trigger words: opencode, coding CLI, delegate coding, multi-provider agent.

First seen Apr 7, 2026

Installation

$ npx skills add wentorai/research-claw --skill opencode-cli

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from wentorai/research-claw · top by installs.

npx skills add wentorai/research-claw

Browse all from wentorai/research-claw

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Declared

Repository health

Stars 851
License LICENSE
Default branch main
Open issues 0
Status Active

Skill metadata

Parsed from SKILL.md frontmatter.

Declared agents claude-code opencode

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 2,935 B
  • docs SUMMARY.md 298 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 12 installs

SKILL.md

OpenCode CLI

Terminal-based AI coding agent supporting multiple LLM providers (OpenAI, Anthropic, Google, local models). Features a TUI for interactive use and run command for automation.

Detect Installation

which opencode  # should return a path
opencode --version

Install

# Recommended (automated)
curl -fsSL https://opencode.ai/install | bash

# npm
npm install -g opencode-ai

# Homebrew (macOS/Linux)
brew install anomalyco/tap/opencode

# pnpm
pnpm install -g opencode-ai

Requires API keys for your chosen LLM provider (configured via /init on first run).

Non-Interactive Usage (run)

For programmatic invocation from Research-Claw, use opencode run:

# Basic non-interactive execution
opencode run "Refactor the auth module to use dependency injection"

# With model selection (provider/model format)
opencode run -m anthropic/claude-sonnet-4-20250514 "Fix all type errors"

# Attach files for context
opencode run -f src/config.ts -f README.md "Update the config schema"

# JSON output for parsing
opencode run --format json "List all exported functions"

# Continue a previous session
opencode run -c "Now add tests for what you just implemented"

# Continue a specific session
opencode run -s <session-id> "Fix the remaining issues"

Key Flags

Flag Description
-m, --model <provider/model> Model in provider/model format
-f, --file <path> Attach file(s) to the message (repeatable)
--format <fmt> Output format: default or json
-c, --continue Continue last session
-s, --session <id> Continue a specific session by ID
--fork Fork session when continuing (new branch)
--share Share the session
--agent <name> Specify which agent to use
--attach <url> Connect to running server (e.g. http://localhost:4096)
--log-level <level> DEBUG / INFO / WARN / ERROR

Research-Claw exec Patterns

# Delegate a complex coding task
exec("opencode run -m anthropic/claude-sonnet-4-20250514 'Implement the feature described in TODO.md'")

# With file attachments for context
exec("opencode run -f README.md -f src/schema.ts 'Add validation based on the schema'")

# JSON output for structured results
exec("opencode run --format json 'List all API endpoints in this project'")

# Using persistent server for multiple tasks
exec("opencode run --attach http://localhost:4096 'Run the test suite and fix failures'")

Docs