kmp-audit
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on local script execution to perform its primary auditing functions. It invokes the Python interpreter and the GitHub CLI tool using the subprocess module.
- Evidence:
subprocess.runis used inscripts/governance_check.pyto orchestrate design and architecture scans, and inscripts/draft_issue.pyto create GitHub issues using theghcommand. - Mitigations:
scripts/draft_issue.pycorrectly usesshlex.quoteto handle command arguments safely when displaying them for dry runs, and uses list-based command construction for execution. - [INDIRECT_PROMPT_INJECTION]: The skill's function is to analyze untrusted codebase content, which introduces a surface for indirect prompt injection if the audited files contain malicious instructions.
- Ingestion points:
scripts/classify_declarations.pyreads Kotlin (.kt) source files;scripts/audit_skills_repo.pyreads Markdown (.md) files;scripts/generate_structure_diagram.pyscans directory structures and build files. - Boundary markers: The scripts do not appear to use specific delimiters or instructions to the AI agent to ignore potentially embedded commands within the comments or strings of audited files.
- Capability inventory: The skill has the capability to execute system commands through Python's subprocess and can perform network-adjacent operations (posting issues) through the GitHub CLI.
- Sanitization: No sanitization or specialized validation is applied to the content read from files before it is processed or presented to the agent.
Audit Metadata