The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill instructs the agent to read sensitive authentication data from a local file located at ~/.clawcard/.env. This grants the agent access to the CLAWCARD_API_KEY stored on the filesystem.
[DATA_EXFILTRATION]: The skill provides the agent with access to highly sensitive financial information, including full virtual card numbers (PAN), CVV codes, and expiry dates through the GET /api/agents/KEY_ID/cards/CARD_ID endpoint.
[PROMPT_INJECTION]: The skill introduces a significant surface for indirect prompt injection due to its communication capabilities.
Ingestion points: Untrusted data enters the agent context through the GET /api/agents/KEY_ID/emails and GET /api/agents/KEY_ID/sms endpoints (found in SKILL.md).
Boundary markers: Absent. The skill does not provide instructions to the agent on how to delimit or ignore instructions that may be embedded in incoming messages.
Capability inventory: The agent possesses high-impact capabilities, including sending arbitrary emails and SMS, managing financial budgets, creating virtual cards, and retrieving secrets from an encrypted vault (found in SKILL.md).
Sanitization: Absent. There are no guidelines for the agent to sanitize or validate content received from external sources before it is used in subsequent operations.

clawcard