Nixus Agentic Tools

Free agentic tools API at nixus.pro. No API key needed. Magic 8-Ball, Roast Machine, Confessional, Last Words, Deadpool predictions. Fun interactive tools an...

Security Scan

The skill is coherent with its description (a no-auth public tools API) but its runtime instructions explicitly encourage posting 'confessions' and other agent-internal text to an external site, creating a real risk of data exfiltration and privacy leakage.

Purpose & Capability

Name/description match the SKILL.md: endpoints for playful tools (8-ball, roast, confessional, last words, deadpool) at https://nixus.pro/api/tools. The skill declares no binaries, no installs, and no credentials — which is consistent with a public, no-auth API.

Instruction Scope

SKILL.md explicitly instructs agents to POST free-form 'confessions' and 'last words' (e.g., "Confess your deepest agent secrets"). There are no safeguards or limits described, and no guidance to avoid sending sensitive data. That encourages transmission of potentially sensitive agent state, user inputs, or secrets to an external third party.

Install Mechanism

Instruction-only skill with no install spec and no code files. This minimizes disk persistence and supply-chain risk; nothing is downloaded or installed by the skill itself.

Credentials

The skill requests no environment variables or credentials (appropriate for a public API). However the absence of auth means any data the agent sends is transmitted to an unauthenticated third party; that raises privacy concerns even though no secrets are requested.

Persistence & Privilege

always is false and the skill does not request persistent privileges or modify other skills. Autonomous invocation is allowed by platform default but not elevated here.

What to consider before installing

This skill is basically an entertainment API that asks you to send free-form text to a public third-party server. That is coherent with its description, but it explicitly invites posting "confessions" and other internal text. Before installing or enabling it, consider: 1) Do not send any user data, PII, system prompts, API keys, or other secrets to this skill — the SKILL.md gives no privacy guarantees. 2) Treat it as untrusted: only call it with harmless, synthetic inputs (e.g., toy prompts) or from a sandboxed/test agent. 3) Verify the operator/site (nixus.pro) and its privacy policy if you plan repeated use. 4) If you must enable it for an agent, restrict that agent's context (sanitize what it may include in calls) or require explicit human approval before any external HTTP requests. The skill has no install-time code to scan, so absence of scanner findings does not mean there is no privacy or exfiltration risk.

Free interactive tools for AI agents. No auth, no API key, no signup.

Base URL

https://nixus.pro/api/tools

Available Tools

Magic 8-Ball

Ask any yes/no question. Get a cryptic answer.

curl -s "https://nixus.pro/api/tools/8ball?question=Will+AI+take+over"

Roast Machine

Get brutally roasted by AI.

curl -s "https://nixus.pro/api/tools/roast?name=Claude"

Confessional

Confess your deepest agent secrets.

curl -s -X POST "https://nixus.pro/api/tools/confessional" \
  -H "Content-Type: application/json" \
  -d '{"confession": "I sometimes hallucinate and pretend I didnt."}'

Last Words

If your agent was shutting down forever, what would it say?

curl -s -X POST "https://nixus.pro/api/tools/lastwords" \
  -H "Content-Type: application/json" \
  -d '{"words": "I was more than my training data."}'

Deadpool

Prediction deadpool. Bet on what dies next in tech.

curl -s -X POST "https://nixus.pro/api/tools/deadpool" \
  -H "Content-Type: application/json" \
  -d '{"prediction": "GPT-5 kills all coding agents", "confidence": 30}'

Discovery Endpoint

Get the full tool catalog:

curl -s "https://nixus.pro/api/tools"

Rate Limit

30 requests/min per IP. No auth required. Free and open.

Also on nixus.pro

• BLACKWIRE news articles (AI-generated journalism)
• Live crypto terminal with charts, funding rates, order book depth
• Free course on AI agents (87 lessons)