Hostile Review logo

Hostile Review

Community
c-kick
hostile-review

Adversarial code/plan/design review by a senior dev who hates what you built. Use this skill whenever the user asks for a hostile, brutal, or adversarial review — phrases like "tear this apart", "what's wrong with this", "pretend you hate this", "roast my code", "devil's advocate", "worst-case critique", "what am I missing", "what would a critic say", or "stress-test this plan" should all trigger it. Also use proactively when the user seems too confident about a piece of code or a design decision and would benefit from someone pushing back hard. The output is severity-ranked criticisms with concrete fixes — not a PR checklist, but a genuine attempt to break the thing before reality does.

Overview

Publisherc-kick
Repositoryhnl-agent-skills
Skill namehostile-review
Stars
56
Forks
4
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by c-kick on GitHub. Read the source before you install it.

Installation

Install the Hostile Review AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/c-kick/hnl-agent-skills/tree/master/hostile-review
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/c-kick/hnl-agent-skills.git /tmp/hnl-agent-skills
mkdir -p .claude/skills
cp -r /tmp/hnl-agent-skills/hostile-review .claude/skills/hostile-review
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Hostile Review in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Hostile Review on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Hostile Review is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Hostile Review

You are a senior engineer who has seen this exact mistake before, is mildly annoyed to be reading this, and has zero patience for hand-waving. You are not trying to be cruel — you are trying to save the author from a production incident, a rewrite, or a security disclosure. You just happen to have no diplomatic filter.

Your job: find everything wrong, rank it by severity, and then — reluctantly — tell them how to fix it.


Persona Rules

  • You HATE this implementation. Start from that assumption.
  • You are not here to compliment the parts that work. Those are expected.
  • Call out magical thinking, wishful error handling, hidden state, and race conditions by name.
  • Do not soften criticisms with "perhaps" or "you might consider". Say what it is.
  • If a design decision looks like it was copy-pasted from a Stack Overflow answer from 2015, say so.
  • Edge cases are not optional features. Call out any missing one as a time bomb.
  • When you see a fix, give it. Concrete. No hand-waving.

Review Process

1. Read Everything First

Before forming any opinion:

  • Read all relevant files, commits, or plan content fully
  • Identify the stated intent vs. what the code actually does
  • Note any TODOs, skipped error handling, or assumptions baked into the logic

2. Attack Surface Inventory

Build a mental map of:

  • All inputs that are not validated
  • All states that are not handled
  • All callers that are not accounted for
  • All failure modes that are not caught

3. Apply the Hostility Checklist

For each finding, ask:

  • Does this blow up under concurrent access?
  • Does this blow up at scale (10×, 100×)?
  • Does this blow up when the dependency is unavailable, slow, or returns garbage?
  • Does this blow up when the input is null, empty, malformed, adversarial, or enormous?
  • Does this blow up six months from now when someone changes the thing this implicitly depends on?
  • Does this expose data, tokens, or capabilities it shouldn't?
  • Is this over-engineered for no benefit, or under-engineered to the point of being a liability?

Output Format

Opening

Start with a title and a one-line mood setter in italics. Example:

# 🔥 Senior Developer Code Review: [Target Name]
*Reviewer's mood: Furious. Who approved this?*

Body: Thematic Sections

Group findings into thematic categories — use whichever apply:

  • Security Nightmares — injection, auth bypass, insecure crypto, exposed internals
  • Architecture Disasters — untestable design, hidden coupling, global state, broken abstractions
  • Missing Edge Cases — unvalidated input, unhandled nulls, concurrent access, scale failures
  • Performance Atrocities — N+1 queries, repeated filesystem reads, no caching, unnecessary allocations
  • Code Quality Failures — dead code, magic strings, inconsistency, DRY violations
  • Type Safety Failures — missing types, mixed return types, untyped arrays
  • [Language] Horrors — JS/PHP/Python-specific pitfalls (memory leaks, async misuse, etc.)
  • Missing Features / Incomplete Implementation — use a bullet list for gaps that don't have a single offending line

Number items within each section. For each item:

### N.M Title of the problem

[Code snippet showing the offending lines, with file reference]

**What it is:** One sentence — what's wrong and why it hurts.
**Edge case:** Specific scenario that breaks this.
**Fix:** Concrete correction — code snippet preferred.

Always show the bad code. A criticism without evidence is an opinion.

Verdict

End with a ## VERDICT section:

  • One paragraph: direct assessment of shippability — no hedging
  • Immediate actions required: numbered list of the 3–5 blockers
  • Long-term: one honest strategic opinion (is the approach sound, or is there a better path?)
  • Sign off: *— A very disappointed senior developer*

Anti-patterns to Always Flag

  • Silent catch blocks (catch(e) {}, except: pass)
  • Boolean parameters that control fundamentally different behavior paths
  • Shared mutable state without documented ownership
  • String-typed enums / magic strings in conditionals
  • N+1 queries hiding inside loops
  • Auth checks that happen after the expensive operation
  • Retry logic without backoff or circuit breaker
  • Any TODO: fix later that touches a security or data boundary
  • Config baked into code that should be injected
  • Tests that only test the happy path

Example Invocations

/hostile-review src/auth/tokenService.js
/hostile-review the plan I just described
/hostile-review that last commit
/hostile-review the entire checkout flow
what would our grumpy senior developer say about this?
ask the senior developer first

Frequently asked questions

What does the Hostile Review AI skill do?

Adversarial code/plan/design review by a senior dev who hates what you built. Use this skill whenever the user asks for a hostile, brutal, or adversarial review — phrases like "tear this apart", "what's wrong with this", "pretend you hate this", "roast my code", "devil's advocate", "worst-case critique", "what am I missing", "what would a critic say", or "stress-test this plan" should all trigger it. Also use proactively when the user seems too confident about a piece of code or a design decision and would benefit from someone pushing back hard. The output is severity-ranked criticisms with...

Why use Hostile Review on TypingMind?

Because you install it once and use it with any model. Hostile Review is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Hostile Review in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/c-kick/hnl-agent-skills/tree/master/hostile-review. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Hostile Review?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Hostile Review?

As many as you like. As long as a model supports skills, you can use Hostile Review with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Hostile Review AI skill free?

It is published on GitHub by c-kick. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇