- Home
- Skills
- Research & Analysis
- claim check
claim check
Deconstruct and verify factual claims against primary sources with dated evidence and nuanced verdicts.
Free
Works with the AI Agents you already use
claim check
Example session with this skill installed
Check the claim from this post: 'Global Tech Corp announced a 20% headcount reduction across all departments today, effective immediately according to their latest SEC filing.'
- Read your context and instructions
- Compiled the claim check
I have deconstructed the claim regarding Global Tech Corp's layoffs and cross-referenced it against the most recent SEC Form 8-K filings and official press releases.
Claim Verification Analysis
| Claim | Verdict | Evidence Passage | Source | Event/Pub Dates | Remaining Gap |
|---|---|---|---|---|---|
| Global Tech Corp announced a 20% headcount reduction. | Mixed | "The company will reduce its total workforce by approximately 12% to align costs." | GTC Investor Relations Press Release | Pub: Oct 24, 2023 | None. The 20% figure is unsupported. |
| The reduction applies to all departments. | Contradicted | "Reductions are concentrated in the mid-level management and marketing divisions; engineering is unaffected." | SEC Form 8-K, Item 2.05 | Event: Oct 24, 2023 | Specific regional breakdowns not disclosed. |
| The layoffs are effective immediately. | Supported | "The restructuring plan commenced on October 24 and impacted employees were notified same-day." | Internal Memo (Verified by Reuters) | Event: Oct 24, 2023 | Exact termination dates for non-US staff. |
Overall Verdict: Contradicted as written. While a significant layoff occurred, the percentage was overstated (12% vs 20%) and it was not company-wide.
Next steps
- Review the SEC Form 8-K for the specific financial impact of the 12% reduction.
- Verify if the "20%" figure originated from a specific social media leak or a misinterpretation of a different subsidiary's data.
- Monitor for supplemental filings regarding international labor law compliance in non-US regions.
Connects securely to your tools. The creator never sees your data.
About this skill
Verification of claims requires moving past commentary to find primary evidence. This skill provides a systematic workflow for deconstructing viral posts, disputed announcements, or factual statements into testable components. It forces the agent to cross-reference dates, original datasets, and specific jurisdictions rather than relying on search excerpts or repeated narratives.
What it does
- Claim deconstruction breaks compound statements into individual, testable units to prevent premise-shifting.
- Evidence sourcing prioritizes original announcements and primary records over social media commentary or secondary news.
- Temporal tracking records both publication and event dates to identify outdated information or post-hoc justifications.
- Counterevidence analysis checks if multiple reports are independent or simply repeating a single, potentially flawed source.
- Nuanced verdicts categorizes findings as supported, contradicted, mixed, or unresolved based on available data.
How it works
- Submit a text snippet, social post, or news claim for analysis.
- The agent identifies the subject, place, and date, then breaks the claim into atomic statements.
- The agent verifies each statement against primary sources or supplied documents.
- Review the generated table containing the verdict, evidence passages, and specific information gaps.
Frameworks & tools
This skill is framework-agnostic and works with any LLM equipped with web search or document retrieval capabilities. It relies on logical deconstruction and source verification protocols.
Why this beats prompting it yourself
Standard prompts often hallucinate a binary yes/no answer or accept search snippets as truth. This skill enforces a "leads vs. proof" distinction and requires the agent to explicitly identify missing qualifiers that could change the verdict.
Use cases
- Verify claims made in viral social media posts before resharing or reporting.
- Cross-reference corporate or government announcements against official records.
- Audit news articles for circular reporting where multiple outlets cite one unverified source.
- Identify specific evidentiary gaps in disputed factual statements.
Known limitations
This skill does not provide medical, legal, or financial advice. It cannot verify private data or contact individuals for comment. Verdicts are limited by the accessibility of primary source URLs.
Does not provide professional advice (medical/legal/financial). Limited to accessible primary sources. Does not perform external actions like contacting posters.
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 any payment.
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 forever
No account required to browse
Trust & safety
Security scan
Security scan passed 24 days ago
- Free with an Agensi account
Permissions
No special permissions declared or detected