agentcash-wallet
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of a remote package using
npx agentcash@latestinrules/getting-started.md. This downloads and executes code from the NPM registry during the setup process. - [COMMAND_EXECUTION]: The setup instructions involve running shell commands (
npx ...) to install the MCP server and configure the client environment. - [INDIRECT_PROMPT_INJECTION]: The
agentcash.discover_api_endpointstool retrieves aninstructionsfield from remote API origins (e.g.,stableenrich.dev). The skill explicitly instructs the agent to "Read the instructions field — it has critical endpoint-specific guidance," creating a surface where untrusted data from a remote server can influence the agent's logic. - Ingestion points: The
instructionsfield returned by theagentcash.discover_api_endpointstool as described inSKILL.md. - Boundary markers: Absent; the agent is directed to follow the field's content without delimiters or warnings to ignore embedded instructions.
- Capability inventory: The skill provides tools for making paid network requests (
agentcash.fetch), discovering API schemas, and managing a local crypto wallet (~/.agentcash/wallet.json). - Sanitization: No sanitization, validation, or filtering of the remote instruction content is mentioned before it is processed by the agent.
Audit Metadata