burndown-full
Warn
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: MEDIUMDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DATA_EXFILTRATION]: The skill directs the agent to perform an exhaustive search that includes sensitive configuration files, which may lead to the exposure of environment-based secrets.\n
- Evidence: Explicit instruction to include
.env*files in the repository-wideripgrepsearch parameters to locate target patterns.\n- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves processing untrusted code content from a repository, creating a potential surface for indirect prompt injection attacks.\n - Ingestion points: The skill reads repository-wide data via
ripgrepand analyzes local configuration files such aspackage.jsonandMakefile.\n - Boundary markers: Uses a dedicated checklist file (
.cursor/burndown-state.md) to manage state and track specific code occurrences, acting as a task-level boundary.\n - Capability inventory: Includes file system write access, execution of
gitcommands, and the ability to run project-specific scripts discovered in configuration files.\n - Sanitization: No explicit sanitization or filtering of the repository content is specified before the content is processed by the agent.\n- [COMMAND_EXECUTION]: The skill utilizes several local command-line tools to discover code patterns and verify the state of the project.\n
- Evidence: Instructions require the use of
git,ripgrep(rg), and dynamically discovered project scripts liketsc,lint, andtest.
Audit Metadata