graph-export
Pass
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
Bashtool to execute a local Python script at.claude/scripts/code_graph. This script is responsible for reading the database and generating the JSON export.\n- [PROMPT_INJECTION]: The skill identifies an attack surface for indirect prompt injection (Category 8) as it processes untrusted source code to generate the graph data.\n - Ingestion points: The skill reads local files (e.g.,
src/auth.py,src/api.py) to build or export nodes and edges.\n - Boundary markers: There are no explicit delimiters or instructions to the agent to ignore potentially malicious content embedded in function names, comments, or class structures within the exported JSON.\n
- Capability inventory: The skill possesses
BashandReadcapabilities, allowing it to execute local scripts and access the filesystem.\n - Sanitization: No explicit sanitization or validation of the extracted code data is mentioned before it is written to the JSON output.
Audit Metadata