task-implementation
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is instructed to search for and execute arbitrary shell commands found in various repository files including
package.json,Makefile, and markdown documentation. - Evidence: The process explicitly commands the agent to 'Run linters' and 'Run tests' by extracting commands from
CLAUDE.md,.cursorrules,AGENTS.md,AGENT.md,GEMINI.md,README.md, andpackage.json. - [PROMPT_INJECTION]: The skill contains instructions that could lead to indirect prompt injection by processing untrusted data from the repository environment.
- Ingestion points: The agent reads
CLAUDE.md,.cursorrules,AGENTS.md,AGENT.md,GEMINI.md,README.md,package.json, andMakefileto find executable logic. - Boundary markers: Absent. No delimiters or instructions are provided to the agent to treat found commands as potentially untrusted or to ignore embedded malicious instructions.
- Capability inventory: The agent has the capability to execute shell commands (subprocess calls) to perform linting and testing.
- Sanitization: Absent. The skill does not include any validation or filtering of the commands extracted from the project files before execution.
Audit Metadata