external-gitcode-ascend-ascendc-operator-doc-writer

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes untrusted data from local source files and external documentation templates without strict safety boundaries. * Ingestion points: The skill reads local C++ source files and reference README documents to extract technical details (SKILL.md, Workflow Step 1). * Boundary markers: No instructions are provided to the agent to use delimiters or to ignore potential commands embedded within the source code comments or the template files. * Capability inventory: The agent has the capability to read local files and communicate with external APIs if requested by the user. * Sanitization: No validation or sanitization of the content read from files is performed.
  • [DATA_EXFILTRATION]: The skill includes a workflow for potential data exposure by instructing the agent to extract source code snippets and send them to external APIs (SKILL.md, Decision Rules). This creates a path for potentially sensitive intellectual property to be sent to untrusted remote endpoints.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 07:21 AM
Security Audit — agent-trust-hub — external-gitcode-ascend-ascendc-operator-doc-writer