The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs users to install the @openfinclaw/cli package from the NPM registry using npx. This is the primary component required for the quantitative analysis and backtesting features.
[COMMAND_EXECUTION]: The skill operates by executing shell commands via the openfinclaw CLI. The provided integration examples for Node.js, Python, and Bash demonstrate how an agent can automate these tools using child_process.exec and subprocess.run.
[CREDENTIALS_SAFE]: The skill encourages secure secret management by advising users to store their API keys in environment variables (OPENFINCLAW_API_KEY) and automatically setting restrictive permissions (chmod 600) on its local configuration file located at ~/.openfinclaw/config.json.
[INDIRECT_PROMPT_INJECTION]: The skill interacts with external financial APIs and community-submitted strategy repositories. While this creates a surface where the agent processes untrusted external data to perform tasks like strategy forking and optimization, this is a standard risk for research tools and is mitigated by the skill's specific focus on financial analysis tasks.

openfinclaw-quantitative-research