Temporal Reasoning Sleuth Skill

The problem

Correlation is frequently mistaken for causation in system logs and project histories. Developers struggle to reconstruct accurate decision chains when timestamps are inconsistent or events are missing from the record.

What it does

• Normalizes multi-source event logs into a single, time-zone-aware chronological timeline.
• Traces decision chains to separate information known at the time from hindsight bias.
• Performs rigorous causality vs. correlation checks using eight distinct logical tests.
• Identifies state changes, turning points, and lag periods across long-horizon event sequences.
• Detects missing events and conflicting evidence between different data sources.

Why this beats prompting it yourself

General LLMs tend to narrativize history, inventing links between events to create a satisfying story. This skill enforces 36 distinct reasoning modes that strictly separate observed facts from causal hypotheses and unknown gaps.

Use cases

• Post-mortem incident reconstruction for distributed system failures.
• Auditing project history to identify where specific technical debt was introduced.
• Analyzing customer journey logs to find the exact trigger for churn or conversion.
• Resolving contradictions between conflicting git logs, meeting notes, and Jira tickets.

Known limitations

Cannot access external systems or live logs; all event data must be provided in the prompt context. It will not speculate on missing timestamps or fabricate events to fill gaps.