qodo-manage-standards
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface as it is designed to ingest and process data from external, potentially untrusted sources such as repository files (e.g.,
CLAUDE.md,AGENTS.md) and pull request history to create or manage review rules. - Ingestion points: Content is read from repository documents and inferred from pull-request comments and history via a rule-mining pipeline.
- Boundary markers: The instructions explicitly require the agent to obtain user confirmation for every write operation and use dry-run modes for bulk actions to prevent unauthorized or accidental modifications.
- Capability inventory: The skill performs subprocess calls using the
qodoCLI to create, update, and delete review rules, and can also perform network operations via the CLI for authentication and catalog management. - Sanitization: The skill relies on structured input schemas and the underlying CLI tool's internal validation, though the instructions do not detail specific sanitization logic for the ingested natural language rules.
Audit Metadata