What problem does it solve? Volumes in the AI-ENGINEERING-OS knowledge base need an objective quality gate before being marked PRONTO, and manual review is inconsistent. This Skill runs a structured audit between gate 2 and gate 3, dispatching a dedicated auditor subagent, recording a dated report, and updating the volume status based on strict scoring criteria. ## Core Features & Use Cases - Subagent-based auditing: Dispatches the auditor-fable subagent with the volume id, section files, dependencies, and a mandatory output format including a media: N.N line. - Machine-readable reports: Saves the auditor output verbatim to auditorias/VOL-NN-auditoria-AAAA-MM-DD.md and verifies the score line is parseable by status.py. - Strict pass criteria: Marks a volume PRONTO only if the average score is at least 8.0, no section scores below 6, and all three validation gates pass; otherwise sets REQUER_REVISAO or RASCUNHO. - Use Case: After finishing a draft volume, run the audit to get an independent scored review, a permanent report in auditorias/, and an automatic status update in _VOLUME.yml and the CHANGELOG. ## Quick Start Audit volume 05 of the acervo by running the gates, dispatching the auditor subagent, and updating its status based on the resulting score.
