Intel Corpus
person concept tool org talk claim — click a node to jump to its page; hover an arrow for the relation
A designated folder within an Agent Workspace that stores an agent's "intelligence corpus" — curated external company knowledge such as meeting notes, emails, and Slack messages that exists outside the codebase or ticket system.
Definition and Purpose
The Intel Corpus is the component of an agent workspace responsible for capturing and preserving organizational reality that lives outside the machine. As Alex Damis describes it, "The Intel corpus remembers the company reality outside of your machine" — distinguishing it from the code itself or the ticket/roadmap system. →
Within the workspace file structure, it is housed in a dedicated corpus folder. →
Architecture and Separation of Concerns
A key design principle advocated by Alex Damis is that the Intel Corpus should be architecturally distinct from other workspace layers (such as the ticket or roadmap system). By treating external company reality — meeting notes, emails, Slack threads — as a separate ingestion layer, the corpus remains more valuable and easier to maintain. This separation ensures that contextual organizational knowledge does not become conflated with task-tracking or project planning structures. →
Components
- LLM Wikis: Noted as a constituent concept within the Intel Corpus, described as a "super cool concept" for organizing LLM-relevant knowledge inside the corpus layer. →
Advocacy
Alex Damis is the primary advocate for the Intel Corpus concept, framing it as essential infrastructure for agents doing long-horizon, "messy forever work" — situations where the agent must maintain awareness of evolving organizational context that no single conversation or codebase can encode. →