The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the BBOT tool via the Python package manager (pipx install bbot), Docker (blacklanternsecurity/bbot), or by cloning the source code from GitHub (github.com/blacklanternsecurity/bbot.git). These references point to the official distribution channels maintained by the vendor, Black Lantern Security.
[COMMAND_EXECUTION]: The core functionality of the skill involves executing shell commands to run the bbot CLI tool. It provides a wide range of examples for reconnaissance tasks such as subdomain enumeration, web spidering, and vulnerability scanning using integrated modules like Nuclei.
[DATA_EXFILTRATION]: The skill documents BBOT's native support for sending scan results to external services, including Discord, Slack, Microsoft Teams, and generic HTTP webhooks. While this is an intended feature for reporting and notifications, it involves the outbound transmission of data gathered during security assessments.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it involves the agent processing and interpreting data retrieved from external, untrusted sources on the internet (e.g., DNS records, web page content, and metadata).
Ingestion points: Untrusted data enters the context via the command-line output of bbot scans performed on target domains and URLs (defined in SKILL.md).
Boundary markers: The instructions do not specify any delimiters or safety warnings to help the agent differentiate between its own instructions and the content retrieved from scanned targets.
Capability inventory: The skill utilizes shell command execution (Category 5) to perform the scans.
Sanitization: No explicit sanitization or filtering of the collected data is described in the instructions.

bbot-recon