tech-debt-tracker
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFECREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: Several files within the
assets/sample_codebase/directory contain hardcoded credentials. While these appear to be placeholder values intended for demonstration purposes (as identifying such debt is the skill's primary function), they represent an unsafe practice in the distributed assets. 1.assets/sample_codebase/src/frontend.js: Contains a hardcodedAPI_KEY("abc123def456"). 2.assets/sample_codebase/src/user_service.py: Contains aDATABASE_URLwith an embedded password ("password123"), a hardcodedAPI_KEY("sk-1234567890abcdef"), and a hardcoded salt ("salt123"). 3.assets/sample_codebase/src/payment_processor.py: Contains several hardcoded keys, includingstripe_key("sk_test_1234567890"),paypal_key("paypal_secret_key_here"), andsquare_key("square_api_key"). - [INDIRECT_PROMPT_INJECTION]: The
debt_scanner.pyscript parses and extracts strings (such as TODO comments and function descriptions) from arbitrary external codebases. This data is then surfaced to the agent, creating a surface for indirect prompt injection where malicious code could attempt to influence the agent's report or subsequent actions. 1. Ingestion points: Source code files in the target codebase directory read byscripts/debt_scanner.py. 2. Boundary markers: The scanner outputs structured JSON and text, but does not provide explicit instructions to the agent to ignore potentially malicious embedded content within the extracted debt descriptions. 3. Capability inventory:scripts/debt_scanner.pyperforms file system reading, AST parsing, and regex pattern matching.scripts/debt_prioritizer.pyandscripts/debt_dashboard.pyprocess the resulting data. 4. Sanitization: There is no evidence of sanitization or filtering of strings extracted from the target codebase before they are included in the generated inventory and reports.
Audit Metadata