graph-blast-radius

Pass

Audited by Gen Agent Trust Hub on Mar 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute a local Python script located at .claude/scripts/code_graph. This script performs operations such as blast radius calculation and dependency tracing by interacting with a local code graph database.
  • [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface (Category 8). Ingestion points: File paths, function names, and class names are ingested from the project's source code and git status. Boundary markers: Absent; code identifiers are interpolated directly into shell command strings without delimiters. Capability inventory: The skill has the ability to execute subprocesses via the Bash tool. Sanitization: Absent; there is no evidence of escaping or validation of these identifiers before they are passed to the shell script.
  • [REMOTE_CODE_EXECUTION]: The skill lists tree-sitter, tree-sitter-language-pack, and networkx as prerequisites. These are well-known and trusted libraries for code parsing and graph analysis. The skill does not attempt to download or execute unverified remote code during operation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 27, 2026, 06:11 AM