Hinsley Documentation
Evidence: Add Your Own Documents
While Hinsley automatically discovers content through its Research Updates, the Evidence tab also holds content that you curate yourself - links, articles, reports, and uploaded documents that you consider important to your analysis. This evidence carries greater weight than standard search results when Hinsley generates scenarios, decompositions, and written outputs.
Adding Evidence
There are several ways to add evidence to an analysis. All methods are available from the Evidence tab in the left sidebar.
Add a URL
Enter a URL in the text field at the top of the Evidence tab. Hinsley will fetch the page content, generate a summary, and assess the relevance of the content to your analysis topic. If the URL has already been added, Hinsley will let you know.
Upload a File
Upload documents directly using the Upload button. Supported file types include PDF, Word (DOCX), plain text, CSV, HTML, and ODT files, up to 10 MB each. Hinsley extracts the text content automatically so it can be used across your analysis.
Bulk Add Links
If you have multiple URLs to add at once, use the Bulk Add Links option. Paste one URL per line and Hinsley will process them all, reporting how many were successfully added, how many already existed, and whether any were invalid.
Save from Search Results
When reviewing search results in the Search Results tab, you can promote any result to evidence by clicking the save icon. This moves the result into your curated collection and gives it the higher priority that saved evidence receives.
How Hinsley Processes Evidence
When you add evidence, Hinsley processes it in the background through several steps:
- Content extraction - For URLs, Hinsley fetches the page and extracts readable content. For uploaded files, it extracts the text from the document format.
- Summarization - Hinsley generates a concise summary of the content for quick reference.
- Relevance assessment - Hinsley evaluates how relevant the content is to your analysis topic, assigning a relevance level of High, Medium, or Low. You can adjust this manually at any time.
Once processing is complete, each evidence document displays its title, source domain, publication date, summary, and relevance level in the Evidence tab.
Relevance Levels
Each evidence document has a relevance level that determines how it is prioritized across your analysis:
- High - Manually added content defaults to High relevance. These documents are prioritized in all AI-generated content.
- Medium - Content that is relevant but not central to the analysis. Still included in generated outputs.
- Low - Tangentially related content. Available for chat and search but excluded from generated outputs.
- None - Setting relevance to None archives the document, removing it from active use.
You can change the relevance level of any evidence document at any time using the dropdown on the document row.
How Evidence Is Used
Evidence is used across Hinsley’s analysis capabilities:
- Scenario generation & likelihood updates - Evidence provides context that shapes how Hinsley generates and refines scenarios and their likelihoods.
- Decomposition - When generating drivers and indicators, Hinsley draws on your evidence to ground its analysis.
- Chat - When discussing your analysis with Hinsley, it can search and read your evidence to provide informed, evidence-based responses.
- Written outputs - Evidence can be included as context when generating action memos, one-page briefs, blog posts, and other outputs.
- Red teaming - You can run a red team analysis against any individual evidence document to stress-test its conclusions.
- Translation - Evidence documents can be translated into other languages if the content supports it.
Evidence Sources
Manually added URLs and uploads remain available independently of the sources selected for automated retrieval. Deselecting a source and choosing Remove collected evidence preserves manual inputs. Existing outputs can still reflect evidence that was removed.