stock-qualitative-analysis

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • REMOTE_CODE_EXECUTION (MEDIUM): The documentation file OPENCODE.md recommends an insecure installation pattern: curl -fsSL https://opencode.ai/install | bash. This pattern is high-risk as it executes unverified remote scripts directly in the shell.
  • COMMAND_EXECUTION (LOW): The script scripts/build_source_manifest.py utilizes subprocess.run to execute fetch_sec_edgar.py. While used for internal logic, this provides a mechanism for system command execution that could be exploited if parameters like tickers or file paths are manipulated.
  • EXTERNAL_DOWNLOADS (LOW): The skill is designed to download financial data from the SEC EDGAR system. While this is its core purpose, interacting with external domains expands the agent's attack surface.
  • PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8) due to its core workflow of processing external, untrusted documents.
  • Ingestion points: scripts/extract_sec_html_sections.py (HTML filings) and scripts/ingest_local_pdfs.py (PDF filings).
  • Boundary markers: Absent. There are no instructions in SKILL.md or the guardrails to wrap external content in delimiters or to ignore embedded instructions.
  • Capability inventory: Use of subprocess.run (in scripts/build_source_manifest.py) and permission to write reports to the local file system.
  • Sanitization: Inadequate. scripts/extract_sec_html_sections.py uses basic regex-based HTML stripping which does not prevent malicious instruction injection hidden within the text content.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 05:41 PM