inboundfy-anti-slop-codigo
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions include the execution of validation tools such as typecheckers, tests, and linters (
SKILL.md). This is a standard procedure for confirming that refactoring efforts maintain functional idempotence and do not introduce errors. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external source code and documentation as its primary input (
SKILL.md). - Ingestion points: The "Entrada esperada" section specifies that the agent receives files, changes, error logs, and refactoring proposals.
- Boundary markers: The skill uses structured YAML frontmatter templates in
REFERENCIA.mdto organize results, which helps differentiate between metadata and the processed content. - Capability inventory: The agent has the capability to write revised code to the filesystem and execute validation commands.
- Sanitization: No explicit sanitization or filtering of input code comments or documentation text is defined in the instructions.
Audit Metadata