The Agent Skills Directory

[PROMPT_INJECTION]: The instruction 'Analyze the mitmproxy dump (log.txt) for OTP issues for: $ARGUMENTS' performs direct interpolation of user-supplied data into the agent's core instructions. This creates a surface for direct prompt injection where a user could provide malicious commands intended to override agent constraints.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from captured network traffic.
Ingestion points: The file log.txt containing external network traffic (referenced in SKILL.md).
Boundary markers: None present; the skill lacks delimiters or instructions to ignore potential commands within the log data.
Capability inventory: The skill utilizes curl for network requests and mitmdump for traffic interception.
Sanitization: There is no evidence of sanitization or validation of the log content before it is processed by the agent.
[COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands including mitmdump, curl requests, and bash loops for testing rate limiting. While these are appropriate for the skill's stated purpose of security auditing, they provide the agent with functional access to the local shell and external network endpoints.

mitm-find-otp