200iq-ai-agent

Enhance your AI with optimized planning, ROI-based decisions, self-correction, and execution for complex tasks and measurable results.

Security Scan

The skill's code is broadly consistent with its stated purpose, but it persistently writes logs to a relative parent directory and the instructions to "log everything" are vague — this creates a risk of writing sensitive data or overwriting files on the host and the delegation language is underspecified.

Purpose & Capability

Name/description match the delivered code: prompt optimization, ROI-based ranking, execution simulation, and self-correction are implemented in the included modules. No unexpected network or cloud credentials are requested, which is consistent with an on-device planning/agent helper.

Instruction Scope

SKILL.md tells the agent to "log everything" and to "delegate sub-tasks to tools/agents" but the code does not constrain what is logged or how delegation is performed. The instructions are vague and grant the agent broad discretion over what to include in logs and what external tools/agents to involve — raising the risk of persisting sensitive inputs or triggering unintended actions.

Install Mechanism

This is instruction-only with embedded Python modules, and there is no install spec or external download. That reduces supply-chain risk.

Credentials

No environment variables, binaries, or credentials are required. The set of declared requirements is minimal and proportionate to the stated functionality.

Persistence & Privilege

logger.py writes to '../logs/journal.log' and '../logs/actions.json' by default (relative path up one directory). Writing to a parent directory can cause persistent files outside the skill bundle, potentially exposing or overwriting host files and persisting sensitive task inputs. log_actions uses 'w' (overwrite) which could clobber existing files. The skill is not marked always:true, but its file-write behavior is a persistent side effect that should be constrained or configurable.

What to consider before installing

The code appears to do what the description claims, but take precautions before installing: 1) Review and change the logger paths so logs are written inside the skill's own sandbox (avoid '../logs' or other relative paths that escape the skill directory). 2) Understand that "log everything" can persist sensitive inputs (passwords, API keys, private data you feed into tasks); avoid sending secrets to the skill or sanitize inputs. 3) If you expect the skill to delegate to other tools/agents, ask for or inspect the implementation of delegation — currently it's only mentioned, not concretely implemented. 4) Run the skill in an isolated environment (container or VM) if you plan to use real data, especially until logging and delegation behaviors are hardened. 5) If you intend to enable autonomous invocation, confirm there are limits on what the agent may call or persist. These steps will reduce the main risks (unexpected persistent writes and accidental logging of sensitive data).

200IQ AI Agent — Genius Mode for Your AI

Turn your AI agent into a 200 IQ genius! Solve complex problems, plan optimally, and execute tasks like a $10M founder combined with a master strategist.

Features

• Decision engine evaluates tasks by ROI and efficiency
• High-IQ prompt optimizer for smarter outputs
• Self-correcting learning loop
• Execution delegation to maximize speed and success
• Action plans with measurable results

Activation

Say:

• "Activate 200IQ AI Agent"
• "Make my AI 200 IQ for this task: [task]"
• "Generate a high-IQ action plan for [goal]"

Usage

1. AI generates optimized prompts and action plan
2. Evaluates which actions produce measurable results
3. Executes tasks with ROI threshold
4. Logs everything and self-corrects
5. Produces outputs the user can implement immediately