The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the longbridge CLI utility across all reference files to retrieve structured financial data, including institutional ratings, consensus targets, and insider trades. These operations are consistent with the stated purpose of the vendor-provided tool.
[EXTERNAL_DOWNLOADS]: For specialized data such as DeFi yields and on-chain metrics, the skill instructs the agent to use WebSearch to fetch information from established financial platforms including DefiLlama, CoinGecko, and Glassnode.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted external data from market news feeds and regulatory filings (as seen in references/stock-research.md and references/hkipo-analysis.md) without utilizing explicit boundary markers or sanitization before processing the content.
Ingestion points: External data enters the context via longbridge news and WebSearch calls.
Boundary markers: Absent.
Capability inventory: The agent can execute shell commands via the longbridge CLI.
Sanitization: No filtering or sanitization of external news or filings is specified.

longbridge-research