The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill documentation instructs the agent to download and execute a shell script from a remote URL using the pattern curl -sSL ... | sh. This is a high-risk distribution method that allows for unverified code execution from the purong-huang-1121/skills-store repository.
[CREDENTIALS_UNSAFE]: The skill documentation explicitly guides users to store their EVM_PRIVATE_KEY in a plaintext .env file. This practice exposes sensitive cryptographic credentials to any local process or user with access to the file system.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted data from the Morpho protocol API. \n
Ingestion points: The plugin-store morpho vault command retrieves metadata.description and metadata.forumLink (SKILL.md). \n
Boundary markers: Absent; there are no instructions to the agent to treat this data as untrusted or to use delimiters. \n
Capability inventory: The skill can execute shell commands via plugin-store and sh (SKILL.md). \n
Sanitization: Absent; the instructions do not specify any validation, filtering, or escaping for the retrieved external text.
[COMMAND_EXECUTION]: The skill relies on executing various shell commands including which, cat, date, and the custom plugin-store CLI to manage its operations and check for updates.

dapp-morpho