The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to interact with the SentiSense API and CLI.
Evidence: Multiple instances of npx -y sentisense@0.47.1 are used to fetch price quotes, sentiment, and insider data. It also provides curl commands for raw HTTP access to the vendor's API.
[EXTERNAL_DOWNLOADS]: The skill downloads and runs a specific version of the vendor's CLI tool from the npm registry.
Evidence: Instructions call for npx -y sentisense@0.47.1. This is a vendor-owned resource required for the skill's primary analysis workflows.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources, which presents a surface for indirect prompt injection.
Ingestion points: Data is fetched from app.sentisense.ai, sec.gov (EDGAR), and fred.stlouisfed.org (FRED) via SKILL.md instructions.
Boundary markers: Uses a highly structured "Evidence Ledger" and a multi-round "Adversarial Committee" pipeline (Steps 0-5) to isolate and verify facts before any synthesis occurs.
Capability inventory: Includes network operations via curl and execution of the sentisense CLI package via npx.
Sanitization: Features a mandatory "Pre-Flight Checklist" to reconcile output numbers against the ledger and a "Fetch safety" section requiring the agent host to implement IP re-resolution, redirect checks, and domain allowlisting.
[CREDENTIALS_UNSAFE]: The skill manages a sensitive API key for its services.
Evidence: It correctly instructs the user to store the SENTISENSE_API_KEY in an environment variable rather than hardcoding it, following security best practices.

stocks-analysis