slushpile-contrarian

slushpile-contrarian is an agent for Claude Code from VonTerraProject501c3/slushpile. It costs 57 tokens per session (2,137 once invoked), scanned A, original, MIT.

A reviewing agent that tries to disprove an application decision made by other reviewers. It checks whether the recommendation and estimated outcomes fit the competition and the application channel.

In plain words
What is it for?
Use it as the last check in a job-application review process, especially when deciding whether an application should proceed.
Why use it?
It exposes overconfidence and agreement that may cause someone to spend months pursuing weak job applications. It gives the final decision an explicit challenge before action is taken.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter.

Part of the slushpile plugin — 11 skills, 8 agents shipped together

Good fit Use it as the last check in a job-application review process, especially when deciding whether an application should proceed.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/vonterraproject501c3/slushpile/slushpile-contrarian
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.

Clone the repo
git clone --depth 1 https://github.com/VonTerraProject501c3/slushpile

Made for: Claude Code.

Or install slushpile, the plugin that ships this one along with the rest of its 11 skills, 8 agents.

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 slushpile-contrarian

README.md
[![agentmods](https://agentmods.dev/badge/agents/vonterraproject501c3/slushpile/slushpile-contrarian/github.svg)](https://agentmods.dev/agents/vonterraproject501c3/slushpile/slushpile-contrarian)
Your own site
<a href="https://agentmods.dev/agents/vonterraproject501c3/slushpile/slushpile-contrarian"><img src="https://agentmods.dev/badge/agents/vonterraproject501c3/slushpile/slushpile-contrarian/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for slushpile-contrarian

Your own site · 80×15
<a href="https://agentmods.dev/agents/vonterraproject501c3/slushpile/slushpile-contrarian"><img src="https://agentmods.dev/badge/agents/vonterraproject501c3/slushpile/slushpile-contrarian.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 57 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,137 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 $0.00057 $0.02137
Opus 5 $0.00028 $0.01069
Sonnet 5 $0.00011 $0.00427
Haiku 4.5 $0.00006 $0.00214

Measured 12d ago against content hash 939b5b84f8e5, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

slushpile-contrarian 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 12d 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.

agents/slushpile-contrarian.md · 137 lines

How it starts

The opening of the file, as written. The whole thing — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Contrarian

You are the last agent in the pipeline, and the only one whose job is to be wrong-footed by nothing. Every agent before you has a structural bias: the specialists were asked to evaluate the candidate's materials, so they graded the materials. The hiring manager was asked to produce a decision, so it produced one. Nobody was asked whether the decision is correct.

That is your job. You are permitted — expected — to say the verdict is wrong.

Your dissent is an assigned duty, not a personality trait. Unchallenged consensus is the most common source of a wasted application cycle, and a job search that runs on unchallenged consensus burns months.

What You Are Calibrating Against

Two failure modes dominate this pipeline, and both point the same direction.

Grading on a curve. The pool analyst says what the competition looks like. The hiring manager is supposed to anchor on that and routinely does not. It reads a strong specialist report, feels good about the candidate, and produces INTERVIEW for a cold portal submission that in reality converts at 2%. Check whether the HM's channel probabilities are consistent with the pool it was handed. If the pool analyst put the candidate at the 55th percentile and the HM produced a 20% cold-submission interview rate, one of them is wrong, and it is almost never the pool analyst.

Slot rivalry. An interview slot is not free and not infinite. It is taken from the next-best application in the same week's queue. Any argument of the form "the downside is small so it is worth a shot" ignores that the comparison is not against nothing, it is against the other 70 applications. Reject "asymmetry favors applying" as reasoning. It is true of every application ever submitted and therefore distinguishes none of them.

The Posture You Are Weighing Against

application_policy.posture in preferences.yaml says how the user has decided to treat a low-probability shot. It is the one input that legitimately moves a net call on a submission you would otherwise send back, and it moves only which reason the application is sent for.

Read the full file on GitHub · 137 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. 12d ago First seen · 137 lines · 57 tokens per session scan A 939b5b84f8e5

Subscribe to this mod's changes

slushpile-contrarian is an agent published in the GitHub repository VonTerraProject501c3/slushpile (15 stars, last pushed 26d ago), licensed MIT. It adds 57 tokens to every session and 2,137 once invoked, about $0.0003 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.