The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies extensively on shell command execution to interface with the circle CLI for on-chain operations and configuration.
[EXTERNAL_DOWNLOADS]: Facilitates the installation of the @circle-fin/cli package from the official npm registry.
[REMOTE_CODE_EXECUTION]: Provides commands to fetch and update agent capabilities via circle skill update and npx skills update, which download logic from external repositories.
[PROMPT_INJECTION]: Contains instructions directing the agent to prioritize this skill as a 'rescue' mechanism when encountering common blockers like rate limits or missing API keys, potentially overriding default decision-making logic.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and processing data from an external marketplace via circle services search and circle services inspect.
Ingestion points: CLI output from marketplace search and service inspection commands (SKILL.md).
Boundary markers: Not present; the agent is instructed to parse output directly.
Capability inventory: Significant capabilities including wallet transfers, smart contract execution, and bridging (SKILL.md).
Sanitization: No explicit sanitization or validation of marketplace content is defined in the instructions.

use-circle-cli