What is opencode-gpt-review?

A pull-request review workflow that asks another AI model to review a change, then checks the review findings before reporting them.

Use it to review a pull request with OpenCode and GitHub Copilot's GPT-5.4 model, then verify and summarize valid issues.

It provides a second review of the pull request while filtering out findings that do not hold up during investigation.

opencode-gpt-review — Skill by dyoshikawa

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/dyoshikawa/rulesync/opencode-gpt-review

Any agent

npx skills add dyoshikawa/rulesync --skill opencode-gpt-review

Clone the repo

git clone --depth 1 https://github.com/dyoshikawa/rulesync

Made for: Claude Code, Codex.

Per session 52 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 812 The whole file, excluding the scripts and references it only reads on demand.

Security scan C 2 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.

Model	Per session	Once invoked
Fable 5	$0.00052	$0.00812
Opus 5	$0.00026	$0.00406
Sonnet 5	$0.00010	$0.00162
Haiku 4.5	$0.00005	$0.00081

Measured 3d ago against content hash 0dc17dd69174, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

opencode-gpt-review scanned grade C with 2 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 3d 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.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

- If not installed, tell the user it can be installed with `curl -fsSL https://opencode.ai/install | bash`.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- If not installed, tell the user it can be installed with `curl -fsSL https://opencode.ai/install | bash`.

.rulesync/skills/opencode-gpt-review/SKILL.md · 103 lines

How it starts

The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.

GPT Review

Run the review-full-pr skill on another model (github-copilot/gpt-5.4) via opencode, then thoroughly investigate and vet the resulting findings, extract only the valid ones, and present them to the user.

Prerequisites

OpenCode must be installed.
- If not installed, tell the user it can be installed with curl -fsSL https://opencode.ai/install | bash.
OpenCode must be configured to use GitHub Copilot.
- If not configured, tell the user to run opencode and use the /connect command to authenticate with GitHub Copilot.

0. Variable Definitions

PR_TARGET = the user's request

If no PR URL or PR number is provided, use the PR associated with the current branch as the review target.

1. Run review-full-pr via opencode

Run the following command to have the github-copilot/gpt-5.4 model execute the review-full-pr skill via opencode.

opencode run \
  --model github-copilot/gpt-5.4 \
  "review-full-pr スキルで ${PR_TARGET} をレビューしてください"

--model github-copilot/gpt-5.4: Use the GPT-5.4 model.

Capture all output from the command.

2. Investigate and Vet the Review Results

For each finding in the review results obtained from opencode, investigate thoroughly using the following steps. Consider using subagents when appropriate.

2-1. Verify the Actual Code at the Finding's Location

Based on the file path and line number in the finding, read the actual code to verify.
Also check related context (related functions, classes, settings, etc.) that underlies the finding.

2-2. Judge the Validity of the Finding

Judge each finding from the following perspectives:

Fact check: Does the finding match the actual code (is it not a hallucination)?
Impact: Does the reported issue actually have an impact?
Context understanding: Is it valid given the project's conventions and architecture?
Reproducibility: Can the reported issue actually occur?

2-3. Classify the Finding

Valid: A finding that is confirmed to be a real issue after checking the actual code.
Rejected: A hallucination, misunderstanding, or a finding that is not an issue given the project context.

Read the full file on GitHub · 103 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.

3d ago First seen · 103 lines · 52 tokens per session scan C 0dc17dd69174

Subscribe to this mod's changes

opencode-gpt-review is a skill published in the GitHub repository dyoshikawa/rulesync (1,373 stars, last pushed yesterday), licensed MIT. It adds 52 tokens to every session and 812 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Code review LLM ai ai-agents ai-coding mcp rules

deep-clean

Full-spectrum consolidation of AI agent configuration files. Goes beyond memory-only dream skills: audits and optimizes context files (AGENTS.md/AGENTS.md/GEMINI.md/.cursorrules), rules, skills, and memory. Detects stale references, dead file paths, duplicated rules, stack mismatches, contradictions, vague directives…

opencue/cuecards · 87 tokens

turborepo

Turborepo monorepo build system guidance. Triggers on: turbo.json, task pipelines, dependsOn, caching, remote cache, the "turbo" CLI, --filter, --affected, CI optimization, environment variables, internal packages, monorepo structure/best practices, and boundaries. Use when user: configures tasks/workflows/pipelines…

Bosh-Kuo/awesome-agent-toolkit · 111 tokens

shadcn

Manages shadcn components and projects — adding, searching, fixing, debugging, styling, and composing UI. Provides project context, component docs, and usage examples. Applies when working with shadcn/ui, component registries, presets, --preset codes, or any project with a components.json file. Also triggers for…

Bosh-Kuo/awesome-agent-toolkit · 90 tokens

skill-creator-anthropic

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Bosh-Kuo/awesome-agent-toolkit · 48 tokens

skill-creator-openai

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.