The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill instructs the agent to read highly sensitive credential files from the local filesystem during the deployment phase, specifically an SSH private key at ~/.ssh/cloudways_rsa and a password file at ~/.cloudways_ssh_password. It also sources project-specific secrets from ~/.pesty-deploy.env.
[COMMAND_EXECUTION]: The skill utilizes several shell commands for environment verification and remote server interaction, including ssh, scp, and curl. This includes executing directory tests and cache purging commands on a remote host.
[DATA_EXFILTRATION]: The deployment workflow in references/deploy.md involves transmitting local project artifacts to a remote server ($PESTY_DEPLOY_SERVER) using the SCP protocol, which constitutes data exfiltration of internal files.
[PROMPT_INJECTION]: The brand audit functionality in references/audit.md creates an attack surface for indirect prompt injection.
Ingestion points: The agent is instructed to fetch and scrape content from arbitrary, user-provided URLs to gather brand data.
Boundary markers: The skill lacks specific delimiters or instructions to treat the scraped content as untrusted or to ignore instructions embedded within it.
Capability inventory: The agent has the ability to read and write to the filesystem and execute shell commands (ssh, scp, curl).
Sanitization: No validation or sanitization of the scraped data is mentioned before it is processed by the agent.

brand-identity