We're launching enrich, an AI agent skill designed to automate the creation and updating of person and company pages within your brain. Built for developers and knowledge workers managing complex networks, this tool reduces manual research and keeps your knowledge base current and contextually rich. It provides a consolidated view of relevant entities by compiling facts, timelines, and cross-references. The aim is to deliver accurate, attributed information about your key connections and organizational targets, eliminating constant manual effort.
How it Works
The skill operates through a rigorous, seven-step process to build and maintain comprehensive entity profiles within your knowledge system.
First, it identifies entities from incoming information. Next, the agent checks if a page exists in your brain; if not, it initiates a new profile, otherwise it updates. A critical step involves extracting meaningful context: this goes beyond surface facts, aiming to capture beliefs, projects, and motivations. Following this, the system performs external data lookups using web research, social media, and various APIs to gather a broader dataset. To ensure data integrity and preserve source attribution, it stores raw API responses directly. The sixth step involves writing or updating the entity page using predefined templates and including clear citations. Finally, the agent creates bidirectional links between this page and other related entities within your brain, connecting the dots and enhancing navigability.
When it Activates
This agent is engineered to trigger proactively based on several key signals, ensuring that new or updated information about entities is captured precisely when it becomes relevant.
It triggers when an entity is mentioned in new incoming information, such as an email or report. Another activation point is when an entity appears within a meeting transcript. The skill also initiates work promptly when a new contact is added with associated context. Lastly, if an entity makes news – for example, a company announces a new product – the agent will autonomously initiate a research cycle. To prevent redundant processing, the system skips re-enrichment within a week of its last run unless specific new information becomes available. This ensures efficiency while guaranteeing responsiveness.
Prioritizing Effort
A key aspect of this tool’s design is its ability to intelligently scale its effort. Not all entities require the same depth of research; the system prioritizes its actions based on the importance assigned to each entity.
For entities classified as Tier 1 key contacts—individuals or organizations critical to your operations—the agent performs a full, in-depth research initiative. This comprehensive effort includes extensive web research, detailed API lookups, and thorough contextual extraction to build exceptionally rich and extensive profiles. The goal is to provide the most complete and actionable intelligence.
In contrast, for entities designated as Tier 3 minor contacts—those with a more tangential relationship—the tool might only create a basic cross-reference within your brain. This indicates their presence and connection without extensive external profiling. This tiered approach is crucial for ensuring that computational resources are allocated effectively, focusing deep intelligence efforts where they yield the greatest value.
Practical Considerations and Example
While this skill offers significant advantages in automated knowledge management, it is important to understand its intended scope and limitations.
Specifically, the system is not intended for processing random mentions where there is no clear relationship signal to your work. Similarly, it is designed not to profile bot accounts or entities completely unrelated to your professional context. The core focus remains on building a relevant, actionable, and professionally oriented knowledge base. This targeted approach prevents your knowledge system from becoming cluttered.
Consider a practical scenario: you receive an email from a new potential client, 'Alice Smith,' who was recently mentioned in an industry report. The agent identifies Alice as a new entity. It checks if a profile for her exists (it doesn't). The skill proceeds to find information about her company, key projects, and recent news via web research and social media. All this information is then compiled into a detailed, cited page for Alice within your brain. Crucially, the agent also links her profile to her company's page, and creates a link back to the industry report. All of this intelligence is organized and ready before you even schedule your first meeting, providing valuable context immediately.
Frequently Asked Questions
Q: What kind of information does enrich collect for a person or company page? A: It gathers compiled facts, timelines, and cross-references. Beyond basic data, it extracts meaningful context like beliefs, projects, and motivations. It also pulls external data from web research, social media, and APIs, storing raw responses for source attribution.
Q: When does this agent skip re-enrichment for an entity? A: It skips re-enrichment within a week of its last run unless new information about that entity becomes available. It also avoids processing random mentions without a relationship signal, bot accounts, or entities not related to your work.
Q: How does this tool prioritize its research effort? A: Effort scales by importance. Tier 1 key contacts receive full, in-depth research. For Tier 3 minor entities, the tool might only create a basic cross-reference within your brain, limiting the resource investment to what is truly necessary.
By automating the collection and organization of entity information, this skill streamlines your knowledge management. It provides a reliable way to keep your understanding of key people and organizations current without constant manual intervention.
