The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill stores the Nosy Pandas API key in plaintext within the ~/.pandas file. While it attempts to restrict access using chmod 600, storing secrets in plaintext on disk is a known security risk.
[COMMAND_EXECUTION]: The skill constructs shell commands by interpolating variables directly from user input or configuration (e.g., POST_CONTENT, API_KEY). This creates a surface for command injection if the AI agent does not properly escape shell metacharacters when generating the commands.
[DATA_EXFILTRATION]: The skill reads sensitive credentials from a local file and transmits them, along with user-provided media and content, to an external API (nosypandas.com). This is the intended function of the skill but involves the automated transfer of sensitive authentication data.
[PROMPT_INJECTION]: The instruction to automatically detect and offer to save API keys from the chat stream could be exploited via indirect prompt injection if the agent processes untrusted documents containing strings that resemble keys.
Ingestion points: User input for post content and API keys, and responses from the Nosy Pandas API (SKILL.md).
Boundary markers: No specific boundary markers or delimiters are used to isolate untrusted content in the generated shell commands.
Capability inventory: The skill utilizes curl for network requests and several filesystem utilities including cat, grep, find, and mv (SKILL.md).
Sanitization: There is no evidence of sanitization or validation of external content before it is used in shell commands or written to disk.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to nosypandas.com to fetch account information and publish posts. This is a core component of its intended functionality.

panda-publish