The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Downloads an installation script for the 'presto' CLI tool from an external domain (https://presto-binaries.tempo.xyz/install.sh).
[REMOTE_CODE_EXECUTION]: Uses the dangerous curl | bash pattern to install the 'presto' utility, which executes remote code without any verification or integrity checks.
[COMMAND_EXECUTION]: Invokes npx to dynamically fetch and execute the awal package, allowing unverified third-party code to run in the environment.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted data from the Codex Supergraph GraphQL API.
Ingestion points: GraphQL query responses from https://graph.codex.io/graphql (found in SKILL.md).
Boundary markers: Absent; the skill does not use delimiters or instructions to ignore embedded commands in the API response.
Capability inventory: Includes shell command execution via curl, npx, and the presto CLI (found in references/wallets.md).
Sanitization: No evidence of sanitization or validation of the data received from the external API before it is processed.

agent-payments