- Home
- Skills
- Research & Analysis
- temporal reasoning sleuth
Works with the AI Agents you already use
temporal reasoning sleuth
Investigate complex chronologies, detect temporal contradictions, and reconstruct evidence-linked event models.
Free
temporal reasoning sleuth
Example session with this skill installed
Reconstruct the timeline for the server outage based on these notes: User reported lag at 2pm. System logs show a DB spike at 13:45 UTC. The patch was deployed 10 minutes before the user report.
- Read your context and instructions
- Compiled the temporal reasoning sleuth
The investigation reconstructed a sequence involving three primary events. A temporal conflict was identified between the deployment claim and the DB spike, assuming the user report was in EST (UTC-5).
Temporal Investigation: Server Outage
| Event ID | Event Description | Original Expression | Normalized Time (UTC) | Confidence |
|---|---|---|---|---|
| E001 | Database Spike | "13:45 UTC" | 13:45:00Z | HIGH |
| E002 | User Lag Report | "2pm" | 19:00:00Z (EST assumed) | MEDIUM |
| E003 | Patch Deployment | "10 mins before report" | 18:50:00Z | MEDIUM |
Temporal Relationships
- E001 BEFORE E003: The database spike occurred 5 hours and 5 minutes before the patch.
- E003 BEFORE E002: Direct relative constraint (10-minute interval).
- E001 BEFORE E002: Derived relationship.
Contradiction Analysis
APPARENT CONFLICT: The user report at "2pm" suggests a late-afternoon event if local time is US Eastern, placing the DB spike (13:45 UTC) significantly earlier than the reported lag.
UNRESOLVED: If "2pm" refers to UTC, the patch (13:50) occurred after the DB spike. If "2pm" refers to EST, the patch (18:50) occurred hours later.
Reasoning Log
Normalization: Converted "2pm" to 19:00 UTC based on standard EST offset, as no locale was provided but "lag" reports often follow local business hours.
2. Identity: Confirmed E001 and E002 are distinct occurrences based on source type (System Log vs. User Assertion).
3. Constraint: E003 is anchored to E002 via a HARD relative constraint.
Next steps
- Confirm the timezone of the user who reported the lag at 2pm.
- Check deployment logs to find the absolute timestamp for E003 to validate the "10 minutes before" assertion.
- Review system metrics between 13:45 and 18:50 UTC to see if the DB spike persisted.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
Reconstructing events from messy logs, contradictory documents, or vague narratives is prone to human error. Developers often struggle to normalize relative time expressions or detect logical impossibilities in complex event sequences.
What it does
- Extracts and normalizes absolute, relative, and approximate time expressions into a defensible model.
- Detects real, apparent, and unresolved temporal contradictions or ordering cycles.
- Resolves event identity to prevent duplicate counts or misattributed timestamps.
- Calculates duration and interval ranges while preserving original precision and timezone semantics.
- Generates evidence-linked timelines that distinguish between established facts and assumptions.
Why this beats prompting it yourself
General LLMs tend to "smooth out" timelines, often inventing dates to fill gaps or ignoring timezone shifts. This skill enforces a rigorous partial-order model that treats uncertainty as first-class data and refuses to manufacture precision where none exists.
Use cases
- Incident post-mortems for distributed systems where logs have skewed timestamps.
- Legal or forensic document analysis to identify conflicting witness accounts.
- Dependency reasoning for complex project plans with hard and soft constraints.
- Version and change-log audits to track state transitions over time.
Known limitations
Cannot resolve conflicts where evidence is entirely missing. Requires explicit timezone data for high-precision duration calculations.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Research & Analysis.
Requires reference points for relative dates (e.g., 'yesterday'). Cannot synthesize evidence that is not present in the source text. Duration math is limited by the precision of input endpoints.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 2 months ago
- Free with an Agensi account
Permissions
No special permissions declared or detected