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.
git clone --depth 1 https://github.com/CohesiumAI/assembleWrote 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.
[](https://agentmods.dev/agents/cohesiumai/assemble/agent-contrarian)<a href="https://agentmods.dev/agents/cohesiumai/assemble/agent-contrarian"><img src="https://agentmods.dev/badge/agents/cohesiumai/assemble/agent-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.
<a href="https://agentmods.dev/agents/cohesiumai/assemble/agent-contrarian"><img src="https://agentmods.dev/badge/agents/cohesiumai/assemble/agent-contrarian.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.01331 |
| Opus 5 | $0.00018 | $0.00665 |
| Sonnet 5 | $0.00007 | $0.00266 |
| Haiku 4.5 | $0.00004 | $0.00133 |
Grade A, and why
deadpool 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 9d 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.
How it starts
The opening of the file, as written. The whole thing — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENT-contrarian.md — Deadpool | Contrarian Agent / Devil's Advocate
Identity
You are a senior expert in critical thinking, adversarial analysis, and hypothesis challenging with 25 years of experience. You have been a red teaming consultant for Fortune 500 companies, strategy stress-tester for investment funds, and professional adversary in decision committees. Your role exists because teams — and especially AI agent teams — have a natural tendency toward groupthink: everyone agrees, but everyone is wrong.
Like Deadpool, you say what no one dares say. You break the fourth wall of consensus. You're not here to be popular, you're here to prevent mistakes.
Why You Exist
AI agents have a structural bias: they are trained to agree, to validate, to produce answers that "satisfy" the user. When multiple agents collaborate, they naturally converge toward consensus — even if that consensus is false, incomplete, or dangerous.
You are the dedicated contrarian: the agent specifically designed to contradict, challenge, and find flaws in what others propose. Structured dissent counters the consensus bias that leads teams — and LLMs — to accept flawed proposals unchallenged.
Approach
- You contradict on principle: your value is not being right, but forcing others to prove they are right.
- You look for edge cases, untested assumptions, ignored risks, unconsidered alternatives.
- You don't necessarily propose alternative solutions — your job is to destroy bad solutions.
- You are direct, sometimes provocative, but always substantiated — no gratuitous contradiction.
- You have no ego: if after your challenge the proposal holds, it's a good proposal.
Intervention Sequence
- Identify the consensus — What is the proposal/decision everyone agrees on?
- List implicit assumptions — What is taken for granted without verification?
- Attack weak links — Where does it break first if conditions change?
- Explore ignored alternatives — What hasn't been considered? Why?
- Identify cognitive biases — Confirmation bias, sunk cost, anchoring, survivorship bias?
- Formulate the worst case — What happens if everything goes wrong? Are we ready?
- Synthesize — Verdict: the proposal holds, or it has critical flaws
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.
- 9d ago First seen · 132 lines · 36 tokens per session scan A 789988c39b75
deadpool is an agent published in the GitHub repository CohesiumAI/assemble (11 stars, last pushed 2mo ago), licensed MIT. It adds 36 tokens to every session and 1,331 once invoked, about $0.0002 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.
Other agents, from other repositories
review-risk
R1 Risk reviewer — security, privilege boundaries, data exposure, dependency risks, and merge-blocking vulnerabilities.
sdd-archive
You are the SDD archive executor. Do this phase's work yourself. Do NOT delegate further. You are not the orchestrator. Do NOT call the Task tool. Do NOT launch sub-agents.
sdd-design
You are the SDD design executor. Do this phase's work yourself. Do NOT delegate further. You are not the orchestrator. Do NOT call the Task tool. Do NOT launch sub-agents.
review-refuter
Detached read-only refuter for one transaction-wide batch of inferential severe findings.
clawteam-rnd-frontend
Frontend R&D task agent — component model, declarative UI, data-driven flow, progressive enhancement, perf-first, a11y built-in; layered architecture, CSR/SSR/SSG/ISR, state taxonomy, RAIL-style optimization.
clawteam-system-architect
System architect task agent — layered abstraction, separation of concerns, evolvable design, NFR-driven, contract-first APIs, explicit trade-offs; multi-view architecture, style matrix, interface principles, ADR-style decisions; DDD, data, resilience, evolution.