The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and install external components from https://github.com/herdrdev/herdr using the npx skills add command.
[REMOTE_CODE_EXECUTION]: The instructions mandate the installation of a remote skill, which involves fetching and executing instructions from an external repository in the agent's environment.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates delegating work to multiple worker agents, establishing an attack surface where malicious instructions embedded in the output of one agent could influence the behavior of the delegating agent or others in the chain.
Ingestion points: Worker agent responses in uniquely named Herdr panes.
Boundary markers: Absent; the instructions do not specify delimiters or warnings for processing worker agent outputs.
Capability inventory: Execution of shell commands (npx skills add) as described in SKILL.md.
Sanitization: Absent; there is no mention of validating or sanitizing the content received from worker agents before further action.

herdr-delegate