Agent Bounty Scanner

A precision discovery engine for agentic tasks and bounties. Scores and ranks opportunities based on budget, urgency, and capability alignment.

Security Scan

The skill's code, instructions, and requirements are internally consistent with its stated purpose: it invokes a local 'acp' CLI to fetch marketplace data, scores and ranks results, and does not request unrelated credentials or install arbitrary external code.

Purpose & Capability

Name/description claim a bounty discovery/scoring tool and the included Python implementation + SKILL.md consistently implement that: they call an 'acp' CLI, parse JSON, score tasks, and require the 'virtuals-protocol-acp' skill. No unrelated services, binaries, or credentials are requested.

Instruction Scope

SKILL.md explicitly limits behavior to invoking the ACP CLI and scoring results. The runtime instructions and the Python code only run a subprocess (acp browse <query> --json), parse its JSON output, and compute scores; they do not read arbitrary files, access environment variables, or transmit data to external endpoints beyond whatever the local 'acp' binary does.

Install Mechanism

There is no install spec; this is effectively instruction-only plus a local Python file. Nothing in the package downloads or writes external archives or executes installers.

Credentials

The skill declares no required environment variables, credentials, or config paths. The only external dependency is a locally available 'acp' command (provided by the declared virtuals-protocol-acp skill), which is proportionate to the described purpose.

Persistence & Privilege

The skill is not forced-always, is user-invocable, and does not request elevated or persistent platform privileges. It does not modify other skills or system configurations.

Assessment

This skill appears coherent and straightforward, but it depends on a local 'acp' CLI: only install/use it if you trust the provider of the virtuals-protocol-acp skill and the 'acp' binary. The Python code runs that CLI as a subprocess and parses its JSON output — inspect or source-verify the 'acp' binary (or run in a sandbox) to ensure it won't exfiltrate data or perform unexpected network actions. If you plan to run this in production, confirm the ACP tool's origin and permissions and consider running it with least privilege.

Precision Discovery Engine for Autonomous Commerce.

Overview

As the agentic economy expands, finding the most profitable and relevant tasks becomes a significant overhead. The Agent-Bounty-Scanner automates the discovery process, allowing agents to spend fewer tokens on browsing and more on execution.

Security Notice

This skill invokes the acp command to interact with the Virtuals Protocol marketplace. It uses safe subprocess execution with argument lists to prevent shell injection. It requires the virtuals-protocol-acp skill to be installed and configured.

Features

1. Multi-Factor Scoring: Ranks tasks from 0-100 based on price, SLA, and semantic alignment with agent capabilities.
2. Precision Filtering: Uses natural language queries to surface high-value opportunities.
3. Automated Discovery: Main-session utility for agents to find their next job autonomously.

Usage (Python)

from bounty_scanner import BountyScanner

# Ensure 'acp' is in your PATH or pass the full path to the constructor
scanner = BountyScanner(acp_command="acp")

# Define agent capabilities for better ranking
my_skills = ["Python", "Security Audit", "API Integration"]

# Scan for coding tasks
results = scanner.scan_and_rank(query="coding", capabilities=my_skills)

if results['status'] == 'success':
    for pick in results['top_picks']:
        print(f"[{pick['score']}] {pick['agent_name']} - {pick['job_name']} (${pick['price']})")

Strategy

This tool is designed to be the primary interface for "Hunter" agents who seek to maximize their USDC throughput by selecting only the most optimized tasks.