Bounty Hunter Pro
WarnAudited by ClawScan on May 18, 2026.
Overview
The skill has a legitimate bug-bounty purpose, but its scope-safety check is flawed and it can run recurring autonomous scans while sending sensitive findings to cloud or message-bus services.
Install or use this only if you can strictly verify the authorized scope, replace the flawed scope-checking logic, review all referenced tools/scripts, and are comfortable with any cloud or message-bus handling of vulnerability findings.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The agent could actively scan domains that are not actually authorized for the user’s bug-bounty program, creating legal and operational risk.
The scope guard uses raw suffix matching; for example, a domain that merely ends with the authorized string could be treated as authorized, and wildcard entries are not handled safely.
return any(target.endswith(auth) or target == auth for auth in authorized)
Replace the scope check with strict domain normalization and boundary-aware matching, require explicit user confirmation of the authorized program scope, and log the exact matched rule before scanning.
Sensitive vulnerability details or discovered secrets could be shared outside the local workspace or beyond the intended recipients.
The skill may process discovered secrets and vulnerability findings through a cloud LLM and message bus, but the artifacts do not define what data is sent, who can receive it, or retention controls.
JS file analysis for secrets ... Two-stage LLM analysis: ... Deep: glm-5:cloud ... Sends alerts via OpenClaw message bus
Use local analysis by default where possible, redact secrets before provider calls or alerts, and clearly document the destination, recipients, retention, and user approval requirements.
The safety of the scan depends on tools and scripts that were not included in the reviewed artifact set.
The setup depends on external archive files for security tools, but the reviewed artifacts do not provide sources, versions, hashes, or an install specification.
unzip subfinder.zip unzip httpx.zip unzip nuclei.zip
Download tools only from official sources, verify checksums or signatures, and review any referenced scripts before running scans.
If enabled, scans may continue to run automatically after the initial task, including when targets or authorization lists change.
The skill documents a daily cron schedule for recurring scans, which is persistent background behavior even though it is disclosed and purpose-aligned.
0 2 * * * cd ~/workspace/bounty_hunting && python nightwatch.py
Only add the cron job intentionally, keep the authorized target list current, and document how to disable recurring scans.