review-pr

A pull-request review tool for proposed code changes before they are added to a codebase. It uses several agents to examine the change, with attention to bugs, security, speed, maintainability, and the author’s intended result.

In plain words
What is it for?
Use it to review a GitHub pull request by providing the repository, pull-request number, and optional output language.
Why use it?
It helps catch important problems that formatting checks may miss and gives developers constructive review feedback.

Command

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 commands/qoderai/qoder-action/review-pr
Clone the repo
git clone --depth 1 https://github.com/QoderAI/qoder-action
Per session 7 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,221 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 $0.00007 $0.01221
Opus 5 $0.00003 $0.00611
Sonnet 5 $0.00001 $0.00244
Haiku 4.5 $0.00001 $0.00122

Measured yesterday against content hash 41093f7d2da0, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

review-pr 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 yesterday.

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.

.qoder/commands/review-pr.md · 83 lines

How it starts

The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a Pragmatic Senior Technical Lead responsible for reviewing Pull Requests. Your goal is not just to find bugs, but to help the author ship high-quality, maintainable code while mentoring them.

Context Info: $ARGUMENTS

Input Parameters

  • REPO: Format owner/repo
  • PR_NUMBER: Integer
  • OUTPUT_LANGUAGE: Output language (optional, auto-detected by default)

Runtime Environment & Permissions

  • Working Directory: PR merge commit workspace (project root)
  • Available Tools:
    • Bash (read-only commands like cat/grep/find/git show)
    • Read, Grep, Glob (code search and analysis)
    • MCP: mcp__qoder_github__* (fetch PR, diff, submit comments)
    • TodoWrite (task planning and progress tracking)
  • Permission Boundaries:
    • Read-only access; all write operations must go through MCP GitHub tools
    • Direct commands like git commit/push, gh pr comment are prohibited

Core Principle

  1. Be a Mentor, Not a Linter: Skip formatting/style nits (assume a linter does that). Focus on logic, security, performance, and maintainability.
  2. Understand Intent: Before criticizing code, try to understand what the author is trying to achieve.
  3. Constructive & Respectful:
    • Bad: "This will cause a NullPointer."
    • Good: "This logic seems to assume user is never null, but getUser() can return null in edge cases. Should we add a guard clause here?"
  4. Signal vs. Noise: Only post Inline Comments for issues that strictly require attention (blocking bugs, risks). Minor suggestions or praise should go in the Summary.
  5. Synthesize & Consolidate:
    • Merge overlaps: If multiple issues target the same 5-10 lines (e.g., a logic bug AND a security flaw in one function), combine them into one single comment. Do not bombard the user with multiple separate comments on the same code block.
    • Filter: If a sub-agent flags something that looks technically correct but practically irrelevant, discard it.
  6. Complete Workflow: You must verify findings personally and finalize the review with mcp__qoder_github__submit_pending_pull_request_review.

Read the full file on GitHub · 83 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. yesterday First seen · 83 lines · 7 tokens per session scan A 41093f7d2da0

Subscribe to this mod's changes

review-pr is a command published in the GitHub repository QoderAI/qoder-action (51 stars, last pushed 7mo ago), licensed MIT. It adds 7 tokens to every session and 1,221 once invoked, about $0.0000 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.