The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes potentially malicious content from external sources (error reports) that could influence agent behavior.
Ingestion points: The skill uses commands like rollbar item and rollbar context to read error titles, messages, and stack traces which are controlled by external users or systems.
Boundary markers: No delimiters or specific instructions are provided to the agent to treat fetched error context as untrusted data.
Capability inventory: The skill has the ability to execute Bash(rollbar *) commands, including status modifications like resolve.
Sanitization: There is no evidence of filtering or sanitizing the data fetched from the external API before it is processed by the agent.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the rollbar CLI tool.
[DATA_EXFILTRATION]: The skill retrieves personally identifiable information (PII), specifically user email addresses and IDs, from the Rollbar error tracking service to provide context for debugging.

rollbar