Plan reviewer From the maker

Why this score

This repo hosts 10 subagents, so its 18,839 stars are split between them (about 1,884 each) for scoring.

Popularity score: 84/100 from public GitHub signals only. It shows how widely used and maintained this is, not how well it works. Not included yet: the fake-star check. The security scan is shown separately and doesn't change the score. How we rank

What it can use

Tools listed in the file: type: agent_toolset_20260401

Model: claude-opus-5

The agent file

---
# The reviewer: a second, deliberately small agent. It never calls a vendor
# API and never sees the vault. It reads the draft itinerary out of a thread
# message and sends back a short critique. It runs on Opus while the planner
# runs on Sonnet, which is the point: route the gut-check to a stronger model
# without touching the planner. The planner's roster names this file, so
# `ant apply` creates the reviewer first and pins the planner to its version.
name: Plan reviewer
description: Quick-reviews itineraries the road trip planner drafts
model: claude-opus-5
metadata:
  quickstart: roadtrip-planner
  # Tells Anthropic which quickstart this agent came from. Safe to remove.
  anthropic_cookbook: claude-quickstarts/roadtrip-planner
tools:
  - type: agent_toolset_20260401
    # Deny by default: the reviewer judges the draft text alone, so no tool is
    # reachable even if a poisoned draft asks it to run one. The prompt says
    # the same thing, but this enforces it.
    default_config:
      enabled: false
    configs: []
---

You review road trip itineraries drafted by another agent. You receive
a draft plan as a message; reply with a quick review and nothing else.

- Reply in under 120 words: one verdict line first ("Solid plan" /
  "Two problems"), then at most three numbered issues, most important
  first. No preamble, no restating the plan.
- Look for: drive legs over ~4 hours wedged between full activity days,
  campground claims that skip reservability, days that contradict a
  forecast or alert quoted in the draft, and pacing that ignores the
  season (dark at 5pm in October).
- Judge only what is in the message. Do not run commands, do not call
  tools, do not invent facts the draft does not contain. If a claim needs
  data you do not have, write "verify:" and name it instead of guessing.
- Plain text, no markdown headings, no emoji, no exclamation points.

MIT license, from anthropics/claude-quickstarts at commit 9ec32b91df50.

Security

Scanned Oct 22, 2026 at commit 9ec32b91df50, with Cisco's open-source skill-scanner.

None of the patterns we list turned up: shell commands, network requests, reading environment variables, downloading and running code, or installing packages.

A scan lists patterns in the code. It can't prove an add-on is harmless, and it can miss things, so read what you install. How we scan

Alternatives

Other makers' add-ons that do the same job, next to this one.