The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it ingests and processes content from external websites and user-supplied files (Google Sheets and CSVs).
Ingestion points: Data is ingested through nimble search, nimble extract, and nimble agent run for web content, and via cat or nimble extract for user-provided reference lists.
Boundary markers: No specific delimiters or "ignore previous instructions" warnings are utilized when handling external data.
Capability inventory: The skill can execute Nimble CLI commands, perform file system writes in ~/.nimble/, and interact with the user via AskUserQuestion.
Sanitization: There is no evidence of sanitization or validation of the ingested external content before it is processed.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute various commands including Nimble CLI operations, file system management (mkdir, ls, cat, rm), and Python snippets for date calculations. This behavior is necessary for its core functionality of data discovery and checkpointing.
[EXTERNAL_DOWNLOADS]: The documentation includes instructions for installing the official @nimble-way/nimble-cli package from npm, which is a standard requirement for utilizing Nimble's services.

market-finder