Full End-to-End Execution Test

[pouria-mirebrahimi]

Story MP-013: Full End-to-End Execution Test

Description:
Run the complete pipeline from ingestion to export for all target symbols and verify reliability, determinism, and correctness across multiple runs.

Tasks:

• Prepare test matrix of target symbols (e.g., AAPL, TSLA, NVDA, AMZN, MSFT) and define 2–3 run profiles (normal, limited-API, retry-scenario)
• Execute full pipeline runs for each profile; capture logs in /logs/pipeline/ and /logs/stages/
• Validate manifest entries creation and content correctness (files list, counts)
• Check determinism: re-run with identical inputs and verify identical outputs (hash/row-count checks)
• Inject controlled failures (e.g., temporary API unavailability) to confirm graceful handling and useful error logs

Acceptance Criteria:

• All target symbols complete E2E runs without unhandled exceptions
• manifest.jsonl entries exist and match exported files and counts
• Deterministic re-run: identical outputs for identical inputs (within allowable timestamp metadata)
• Failure scenarios produce structured error logs and recover cleanly on next run

Estimation: 2.5 days
Dependencies: MP-012
Status: Planned

[pouria-mirebrahimi]

Scripts and CLI commands

@nikbakhtenb4