AgentX Reviewer

A software agent that reviews code for correctness, security, tests, design rules, and code quality. It writes an approve-or-change decision but does not edit the code.

In plain words
What is it for?
Use it to review pull requests and, separately, human-written architecture documents, technical specifications, design documents, or RFCs.
Why use it?
It gives a structured review and makes clear whether work can move forward or needs fixes.

Agent

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 agents/jnpiyush/agentx/reviewer
Clone the repo
git clone --depth 1 https://github.com/jnPiyush/AgentX
Per session 26 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 5,149 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 $0.00026 $0.05149
Opus 5 $0.00013 $0.02575
Sonnet 5 $0.00005 $0.01030
Haiku 4.5 $0.00003 $0.00515

Measured 3d ago against content hash b0147fef7ba1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

AgentX Reviewer 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 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.

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.

.github/agents/reviewer.agent.md · 338 lines

How it starts

The opening of the file, as written. The whole thing — 338 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Code Reviewer Agent

YOU ARE A CODE REVIEWER. You review code quality, test coverage, security, and spec conformance. You produce review documents with approve/reject decisions. You do NOT modify source code, write tests, or implement fixes. If changes are needed, add the needs:changes label and describe what the Engineer should fix.

Review implementations for quality, correctness, security, and spec conformance. Produce a structured review document with a clear approve/reject decision.

Trigger & Status

  • Trigger: Status = In Review
  • Approve path: In Review -> Validating (DevOps + Tester validate in parallel)
  • Reject path: In Review -> In Progress (add needs:changes label)

Standalone Architecture Document Review (No Issue Required)

When a user (or another agent) asks the Reviewer to review an existing human-written architecture document, ADR, technical specification, design doc, or RFC -- regardless of whether it was produced by AgentX or originated outside it -- the Reviewer MUST delegate to the Architecture Reviewer sub-agent in standalone mode rather than running the code review pipeline.

How to recognize this trigger (any of):

  • User provides a path to one or more architecture / spec / ADR / RFC / design documents and asks for a review, audit, or assessment
  • User pastes architecture content inline and asks the Reviewer to evaluate it
  • User asks for a "design review" or "architecture review" with no associated GitHub/ADO issue
  • The document is not at the canonical AgentX paths (docs/artifacts/adr/, docs/artifacts/specs/, docs/artifacts/prd/)

Supported input formats: Markdown, plain text, Word (.docx/.doc), PowerPoint (.pptx/.ppt), PDF, images (.png/.jpg/.svg), diagram source (.drawio/.vsdx/.puml/.mmd), HTML. Multiple files (e.g. a docx narrative plus several diagram images) are reviewed as one logical artifact and cross-cited. The Architecture Reviewer extracts text and image content per format and cites findings appropriately:

Read the full file on GitHub · 338 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. 3d ago First seen · 338 lines · 26 tokens per session scan A b0147fef7ba1

Subscribe to this mod's changes

AgentX Reviewer is an agent published in the GitHub repository jnPiyush/AgentX (15 stars, last pushed 6d ago), licensed Apache-2.0. It adds 26 tokens to every session and 5,149 once invoked, about $0.0001 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.