opencontext

Warn

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to install a global NPM package @aicontextlab/cli and download software releases from a GitHub repository (0xranx/OpenContext). These sources are not part of the trusted vendors list.
  • [COMMAND_EXECUTION]: The skill relies on the execution of CLI commands like oc init and oc search. It also mentions running arbitrary shell commands, such as docker-compose up -d, within a multi-agent workflow.
  • [CREDENTIALS_UNSAFE]: The skill documentation includes instructions for users to save sensitive credentials, specifically an EMBEDDING_API_KEY, using the tool's configuration command.
  • [PROMPT_INJECTION]: The skill processes data from a local knowledge base, creating a potential surface for indirect prompt injection.
  • Ingestion points: Untrusted data enters the context through oc search and oc_list_docs as described in SKILL.md.
  • Boundary markers: There are no specific delimiters or instructions to ignore embedded commands within the retrieved documents.
  • Capability inventory: The skill performs subprocess calls via shell, file writing via oc doc create, and network operations for API access.
  • Sanitization: No sanitization or validation of the ingested content is documented.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 7, 2026, 07:03 AM