The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The library provides functionality to save and load filing data using the Python pickle module (e.g., Filing.save and Filing.load in references/filings.md). Deserializing data from untrusted .pkl files is a known vector for arbitrary code execution.
[PROMPT_INJECTION]: The skill processes external SEC filing content which creates a surface for indirect prompt injection attacks.
Ingestion points: Functions such as filing.markdown(), filing.text(), and filing.html() in references/filings.md ingest external document content into the agent context.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore commands embedded in the retrieved data.
Capability inventory: The agent is granted capabilities including file system writes, network access to SEC servers, and Python code execution.
Sanitization: There is no documentation regarding the sanitization or validation of content retrieved from the SEC.
[DATA_EXFILTRATION]: The skill requires the user to provide their name and email address via set_identity or the EDGAR_IDENTITY environment variable. This identity information is transmitted as a header to SEC servers with every request to satisfy regulatory access requirements.
[COMMAND_EXECUTION]: The documentation includes shell commands for installation and server configuration, such as uv pip install, uvx, and python -m edgar.ai. Additionally, the install_skill() function in references/ai-integration.md writes files to the user's ~/.claude/skills/ directory for persistence.
[EXTERNAL_DOWNLOADS]: The skill is designed to programmatically fetch and download large-scale financial datasets and regulatory filings from the official SEC EDGAR system.

edgartools