The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructs the agent to download a script from https://raw.githubusercontent.com/codetesla51/barrage/main/install.sh and pipe it directly to bash, which executes code from a repository not associated with a trusted organization.
[EXTERNAL_DOWNLOADS]: The skill fetches installation scripts and Go packages from the codetesla51/barrage GitHub repository, which is external to the skill author (phenpen) and not a recognized trusted vendor.
[COMMAND_EXECUTION]: The skill requires the execution of multiple shell commands, including go build, go run, and the barrage CLI tool itself, which can pose a risk if the source repository is compromised.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external HTTP endpoints, databases, and Redis instances, creating an attack surface for indirect prompt injection via the load testing results.
Ingestion points: The http, db, and redis runners ingest data from target systems as defined in the barrage config files.
Capability inventory: The skill executes shell commands using the barrage binary and Go compiler tools.
Boundary markers: There are no explicit instructions to use delimiters or ignore embedded prompts in the data being analyzed.
Sanitization: The instructions do not specify any sanitization or filtering of the content retrieved from external sources during testing.

barrage