Start with the project outcome
Define what the project should achieve and which assumptions remain unresolved. Cohervio investigates a business request by examining the people, systems and open questions behind it. Its process moves from understanding the request through discovery, interviews, investigation, specification, and alignment and approval.
Turn interviews into traceable material
Keep each interview note with its source. Cohervio links every requirement to the interview, document or system behind it, and returns an evidence-linked specification plus the decisions still needed. When specifying reveals a missing assumption, the investigation can loop back to the right stakeholder.
Capture meeting minutes and decisions
Save the meeting link and add a note that records the decision, owner and open question. KnowMan saves articles and links, reads and tags them, and keeps memories such as claims, decisions and playbooks. This gives your project notes a shared home instead of separate documents and chat threads.
Organize the specification record
Group material by project, workstream or decision type. For the specification itself, Cohervio includes a stakeholder map showing who owns each answer, approvals showing who signed off on what, and a risk report listing items that still need validation. Keep unresolved items visible rather than treating them as settled.
Search by meaning
Search for ideas, not only exact titles. KnowMan searches saved items by keyword or meaning across titles, URLs and metadata. This helps you find the relevant interview note or decision when you remember the point rather than the document name. See the guide to searching a personal knowledge base.
Ask questions with citations
Ask a focused question when you need the current state of the project. KnowMan answers questions with citations to the saved source material. You can also read the full content of one saved item. Learn more in the guide to giving an AI context.
Connect an AI coding agent
KnowMan provides an MCP server for AI agents. It runs as a local stdio server with npx -y @knowman/mcp and Node 22 or newer, and as a hosted remote server at https://api.knowman.io/mcp over Streamable HTTP. Its tools include kb_search, kb_ask, kb_get_item, kb_save, kb_recent and kb_memories. See the guide to MCP server knowledge bases.
Where KnowMan fits
KnowMan serves as the searchable library for interviews, meeting minutes, decisions and supporting links. It can also supply an AI coding agent with that material through the MCP server. The setup steps for connected clients are covered in Claude Code knowledge bases and Claude knowledge bases.
Cohervio explains the specification side in feed your coding agent the whole picture.
Questions people ask
What belongs in a specification knowledge base?
Start with the business outcome, then gather interviews, supporting documents, system constraints and open questions. Cohervio records what stakeholders said, what existing systems can support and what remains open, so a team can see what is known and what still needs a decision.
How do I keep interviews, minutes and decisions together?
Save the links and articles in KnowMan, then add notes that identify the source, decision and owner. KnowMan saves articles and links and keeps memories such as claims, decisions and playbooks, so related material stays in one library.
Can an AI coding agent read the specification library?
Yes, through the KnowMan MCP server, which can run locally over stdio or as a hosted remote server. Its tools support search, cited answers, reading one item and retrieving recent items. See MCP server knowledge bases for the setup.
Does KnowMan connect directly to Cohervio?
There is no direct integration described here. The two serve different stages of the work, with KnowMan holding the saved research and Cohervio producing the specification. Keep the specification and its source links in the same KnowMan library.