The Agent Skills Directory

[DATA_EXFILTRATION]: The skill contains instructions to execute commands that print sensitive environment variables into the agent's execution context for debugging.
Evidence: echo $INSTANTLY_API_KEY and env | grep INSTANTLY in SKILL.md are used to verify local configuration, which exposes the content of these secrets to the model.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests untrusted data from external sources (logs and console output).
Ingestion points: Error messages and codes retrieved from user logs or console output (Step 1 in SKILL.md).
Boundary markers: No boundary markers or instructions to ignore embedded commands within the logs are provided.
Capability inventory: The skill has access to Bash (specifically curl), Grep, and Read tools.
Sanitization: No sanitization, escaping, or validation of the log content is performed before processing.

instantly-common-errors