dev-context-code-graph

Pass

Audited by Gen Agent Trust Hub on Oct 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a repository analysis pipeline using standard Python libraries and deterministic parsing (AST for Python, Regex for others). No malicious behavior, obfuscation, or unauthorized network activity was detected.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external source code through scripts/scan_code_repo.py. While this creates an inherent surface for indirect prompt injection if the scanned code contains malicious instructions, the risk is minimal as the output is structured JSON and the behavior is central to the skill's primary function. Sanitization (HTML escaping) is present in reporting scripts.
  • Ingestion points: scripts/scan_code_repo.py reads repository file contents.
  • Boundary markers: None explicitly defined in the processing scripts.
  • Capability inventory: File read/write operations for graph and report generation; no network or subprocess capabilities are exposed to the processed data.
  • Sanitization: Basic HTML entity escaping is performed in scripts/export_code_graph_report.py.
  • [DYNAMIC_EXECUTION]: The test suite in scripts/test_code_graph_regressions.py uses importlib.util for module loading of local scripts during testing. This implementation uses static paths relative to the test script and is a standard development practice.
  • [EXTERNAL_DOWNLOADS]: The skill references several external sources in data/sources.json, including official documentation for Python, JSON Schema, and Sourcegraph, as well as reference implementations on GitHub. All referenced domains and organizations (GitHub, Sourcegraph, Python.org, JSON-Schema.org) are well-known services.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 6, 2026, 02:03 PM