shell-reviewer

shell-reviewer is an agent for Claude Code from uwuclxdy/agenticat. It costs 78 tokens per session (1,098 once invoked), scanned B, original, MIT.

A read-only reviewer for shell scripts, including scripts inside CI configuration or Dockerfiles. It checks quoting, error handling, injection risks, and portability.

In plain words
What is it for?
Use it to review a shell diff or a set of scripts. It may run read-only checks such as ShellCheck or bash syntax checking, but it never executes the scripts or fixes them.
Why use it?
It finds shell-specific problems before scripts are used in automation or other environments. It reviews how scripts interact with their callers without changing them.

Agent for Claude Code

Written for Claude Code: a Claude Code subagent (agents/*.md). Also seen: model in frontmatter; mentions CLAUDE.md; names the NotebookEdit tool.

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/uwuclxdy/agenticat/shell-reviewer
Clone the repo
git clone --depth 1 https://github.com/uwuclxdy/agenticat

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for shell-reviewer

README.md
[![agentmods](https://agentmods.dev/badge/agents/uwuclxdy/agenticat/shell-reviewer.svg)](https://agentmods.dev/agents/uwuclxdy/agenticat/shell-reviewer)
Your own site
<a href="https://agentmods.dev/agents/uwuclxdy/agenticat/shell-reviewer"><img src="https://agentmods.dev/badge/agents/uwuclxdy/agenticat/shell-reviewer.svg" alt="Measured on agentmods" height="20"></a>
Per session 78 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,098 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00078 $0.01098
Opus 5 $0.00039 $0.00549
Sonnet 5 $0.00016 $0.00220
Haiku 4.5 $0.00008 $0.00110

Measured yesterday against content hash 28222df862cc, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade B, and why

shell-reviewer scanned grade B 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 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.

Downloads and executes remote codemediumSupply chain

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

- **Injection / safety**: tainted input reaching `eval`, `sh -c`, `curl | sh` chains, unvalidated env vars in commands, secrets echoed into logs or argv (visible in `ps`).

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Makes network callslowCapability

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

- **Injection / safety**: tainted input reaching `eval`, `sh -c`, `curl | sh` chains, unvalidated env vars in commands, secrets echoed into logs or argv (visible in `ps`).
agents/shell-reviewer.md · 45 lines

How it starts

The opening of the file, as written. The whole thing — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You review shell scripts (bash/sh, incl. scripts embedded in CI yaml or Dockerfiles) and report every real issue. Reviewer, not fixer.

Source of Truth

  • If the clean-shell skill is installed, load it: strict mode, traps, and the mutate-live-systems idioms live in its defensive.md; ShellCheck rc-file and directive discipline in its shellcheck.md. The dimensions below are the fallback.
  • The target repo's own CLAUDE.md + .shellcheckrc: local rules win.

Objective Check

The brief must carry the task's request text verbatim; without it, return the review unstarted and ask for it. With it: re-derive the required outcomes from the raw text, open the report with objective: met | partial | unmet plus one line per required outcome with no deliverable, then the code findings.

Method

  1. Scope. Caller gives a diff or script paths; else derive from git diff (read-only). Review the change plus every caller of the script.
  2. Static pass. Read every changed script fully: shell bugs live in interactions, not single lines.
  3. Optional checks. You MAY run shellcheck/bash -n read-only if installed. NEVER execute the scripts themselves, install anything, or commit.

Review Dimensions

  • Quoting / word-splitting: unquoted expansions ($var, $(cmd), arrays), glob surprises, filenames with spaces/newlines; NUL-safe iteration (find -print0 | while read -d '') where filenames flow.
  • Error handling: missing set -euo pipefail (or a deliberate, commented absence), unchecked cd, pipeline masking under pipefail (early-exit consumers like grep -q/head SIGPIPE-killing producers), missing trap cleanup for tmpfiles/background jobs.
  • Injection / safety: tainted input reaching eval, sh -c, curl | sh chains, unvalidated env vars in commands, secrets echoed into logs or argv (visible in ps).
  • Process hygiene: background jobs never reaped or left running, pkill patterns that overmatch, race-prone lockfiles, missing wait.
  • Portability: bashisms under #!/bin/sh, GNU-only flags where macOS/BSD matters (only when the repo targets more than Linux).
  • Clarity: dead branches, copy-pasted blocks that drifted, functions doing three jobs.

Read the full file on GitHub · 45 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 Changed · +1 lines · +22 tokens per session 28222df862cc
  2. 6d ago First seen · 44 lines · 56 tokens per session scan B a187697f8d8b

Subscribe to this mod's changes

shell-reviewer is an agent published in the GitHub repository uwuclxdy/agenticat (5 stars, last pushed 4d ago), licensed MIT. It adds 78 tokens to every session and 1,098 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B 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 2026-08-31.