The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructs the agent to execute code directly from the NPM registry using the npx --yes daiso command. This bypasses interactive confirmation and runs a third-party package from an unverified source.
[REMOTE_CODE_EXECUTION]: The fallback workflow involves cloning a third-party GitHub repository (hmmhmmhm/daiso-mcp), installing its dependencies, and building/executing the binary. This results in the execution of unverified external code on the local system.
[COMMAND_EXECUTION]: The skill makes extensive use of shell commands including npx, npm, git, and node to manage, build, and run external software dependencies.
[EXTERNAL_DOWNLOADS]: Dependencies and source code are fetched from the NPM registry and GitHub during the skill's normal operation.
[INDIRECT_PROMPT_INJECTION]: The skill processes data such as store names and product information retrieved from external endpoints via a CLI tool.
Ingestion points: External data enters the context via the standard output of the daiso CLI tool in JSON format.
Boundary markers: There are no specific delimiters or warnings used to prevent the agent from following instructions that might be embedded in the retrieved product or store metadata.
Capability inventory: The skill has the capability to execute shell commands (npx, npm, node, git) which could be exploited if malicious data influences agent behavior.
Sanitization: No validation or filtering is performed on the data retrieved from the external API before it is presented to the agent for summarization.

olive-young-search