- Home
- Skills
- Research & Analysis
- Reduce to Facts
Reduce to Facts
Extract a source-aware fact ledger from dense or ambiguous workplace material before replying, deciding, or acting.
Free
Works with the AI tools you already use
Reduce to Facts
Example session with this skill installed
Analyze this thread regarding the Project Alpha delay. Multiple leads are blaming different teams. I need to know what is actually confirmed versus what is just finger-pointing.
- Read your context and instructions
- Compiled the reduce to facts
- Generated the document
I have reviewed the Project Alpha thread. I have extracted the atomic facts and separated them from team positions and unsupported claims to clarify the source of the delay.
Source Scope
- Source(s) reviewed: Project-Alpha-Delay-Thread.txt
- Source type: Slack Export / Email Chain
- Time period covered: Oct 12 - Oct 15, 2023
- Missing context: Response from the DevOps lead regarding the server migration timestamp.
Atomic Facts
| ID | Fact | Source anchor | Confidence | Notes |
|---|---|---|---|---|
| F01 | Frontend build failed at 09:00 UTC | @jane-doe (Oct 12) | High | Confirmed by build logs attached to post |
| F02 | API credentials were rotated on Oct 11 | @ops-team (Oct 11) | High | Direct system notification cited |
| F03 | 12 tickets remain in the "In Progress" column | Jira Link 402 | Medium | Status may have changed since report |
Attributed Opinions / Positions
| ID | Who/Source | Position | Basis | Confidence |
|---|---|---|---|---|
| O01 | @jane-doe | Delay is due to API auth changes | Failure started after credential rotation | Medium |
| O02 | @backend-lead | Delay is due to inefficient CSS bundling | Build time increased by 400% | Low |
Unsupported or Under-Supported Claims
| ID | Claim | Why unsupported | Evidence needed |
|---|---|---|---|
| U01 | The database was down for 2 hours | No logs or status alerts support this | Database heartbeat logs |
Inferences
| ID | Inference | Based on | Risk if wrong |
|---|---|---|---|
| I01 | The rotation caused the build failure | F01, F02 | High - may overlook underlying CSS issues |
Contradictions / Tensions
| ID | Conflict | Sources involved | Needs resolution |
|---|---|---|---|
| C01 | Auth vs. Bundling as root cause | @jane-doe vs @backend-lead | Yes - requires log deep dive |
Next steps
- Use
decision-briefto choose between roll-back or hotfix. - Use
clear-askto get the DevOps lead's specific migration timeline.
reduce-to-facts.pdf
PDF · document
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Stop making high-stakes decisions based on loose summaries or messy Slack threads. This skill extracts objective truth from dense, contradictory, or stakeholder-sensitive workplace materials before you act. It prevents you from acting on assumptions by creating a rigorous fact ledger that separates stated facts from opinions and inferences.
What it does
- Atomic fact extraction creates a ledger of individual, verifiable claims mapped to specific source anchors.
- Uncertainty preservation identifies where sources are ambiguous or missing context rather than smoothing over gaps.
- Confidence scoring assigns High, Medium, or Low values to every claim based on directness and source strength.
- Contradiction mapping flags hard conflicts and soft tensions between different participants or documents.
- Source-expanded verification fact-checks internal claims against outside evidence only when explicitly requested.
- Rhetorical analysis distinguishes between what was actually said and the broader implications or leaps in reasoning.
How it works
- Submit source material such as email chains, strategy memos, or customer escalations.
- Review the Fact Ledger to see atomic facts, attributed opinions, and unsupported claims.
- Verify implications by reading the inference table which links logical conclusions to specific data points.
- Identify blockers using the open questions and contradiction tables to resolve issues before proceeding.
Frameworks & tools
Works with any text-based workplace source material including Slack exports, Zoom transcripts, PDF policy documents, and markdown memos.
Why this beats prompting it yourself
Generic summaries often "hallucinate" consensus where there is actually conflict. This skill uses a structured confidence rubric and specialized tables to ensure you don't mistake a stakeholder's opinion for a hard technical constraint.
Use cases
- Auditing a long customer escalation thread to find the technical root cause versus the emotional complaints.
- Analyzing contradictory vendor writeups to identify specific gaps in a proposed solution.
- Preparing for a high-stakes meeting by mapping out all stakeholder positions and open questions from prior memos.
Known limitations
Does not perform external research unless source-expanded mode is triggered. Output length increases significantly with source density.
How to install
Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.
- 1
Download the ZIP
Free skills download straight away. Paid skills unlock right after purchase.
- 2
Unzip into your skills folder
Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.
- 3
Ask your agent to use it
Restart the agent if it was already running. It picks the skill up automatically - no config needed.
Skills folder by agent
Click the path to copy it. Create the folder if it does not exist yet.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scanned
Verified clean 1 day ago
- Free to download with an account