audit-and-format

Automating the audit and formatting phase for a stock recommender pipeline requires executing closed-loop learning steps in strict order, including self-audit, miss audit, and rule ledger updates to ensure data integrity. This skill performs those exact tasks.

Updating a rule ledger with upvote and downvote status recomputation involves processing pipeline audit results and appending them to JSON files. This skill automates the rule ledger update by reading and writing to multiple JSON and CSV files.

Appending daily stock recommendations to a JSON file requires a structured data append process after completing pipeline audits. This skill automates appending daily recommendations directly to daily_recommendations.json as part of its formatting phase.

Running a self-audit and miss audit using Bash commands alongside JSON and CSV files is fully supported. The skill executes these audit tasks by reading from and writing to multiple data files while invoking Bash commands.

Token cost reports in a data processing pipeline are needed during the audit phase to track computational expenses. This skill generates token cost reports as part of its strict closed-loop learning step execution sequence.

Limitations include its strict dependency on the india-stock-recommender pipeline structure, requiring specific JSON and CSV file formats. It is not a general-purpose data processing tool and is tailored for closed-loop learning audit tasks.