The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs the nansen-cli package via NPM during the setup process to enable its research functionality. Nansen is a well-known blockchain analytics platform.
[COMMAND_EXECUTION]: The skill executes nansen CLI commands through the Bash tool. The allowed-tools metadata restricts execution to the nansen command prefix, which limits the risk of executing unauthorized system commands.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from the Nansen API (blockchain addresses and labels) and uses these values as variables for follow-up research commands, creating a surface where malicious data could influence agent actions.
Ingestion points: Token addresses and trader labels retrieved from the nansen research smart-money dcas command in SKILL.md.
Boundary markers: The skill does not employ specific delimiters or "ignore previous instructions" warnings when processing the API output.
Capability inventory: The agent has the ability to run scoped bash commands (nansen:*) and interact with external network services via the Nansen CLI.
Sanitization: There is no explicit validation or sanitization of the address strings extracted from the API before they are interpolated into shell commands.

nansen-dca-tracker