How it's different

Six points the owner set, not a feature checklist against a named competitor. Each is set beside what a team reaches for today, by kind of tool rather than by name.

onMainIssue tracker (sprints)A wikiMarkdown in the repo
No Agile ceremonyIdea → stream → card, each pinned to a plan.Initiative, epic, story, task — sprint-planned and estimated.No tracked unit of work at all.A file per decision at best — nothing queued or ordered.
Built for agentsEvery write path is an MCP tool call.A UI-first API, fitted on after the fact.A page-editing API at best, not shaped for an agent’s tool calls.No API — an agent edits the file the way a person does.
Three loopsA published, pinned process read before every plan.Workflow states and transitions, not a process document.A process page nothing requires anyone to read or keep current.A CONTRIBUTING.md or similar — versioned by git, but pinned and enforced by nobody.
The PR Train and the PR ResolverEvery ready pull request is merged in order by the train; the Resolver, which attempts the conflicts first, is coming.Not what a tracker does — merging is the repository’s job.Not what a wiki does.Not on its own — merging is still by hand.
The Knowledge Hub and graphOne connected graph of every record, walkable by a person or an agent — coming.Linked issues at best, no decision or requirement type.Pages that cite each other by hyperlink, with no enforced structure.Files that reference each other by path; nothing checks the links hold.
Records that do not driftA decision names what it rests on and what would overturn it; a plan takes it; status follows approval.A card can be labelled or linked, not structured as a decision.A decision log page, free text, with nothing stopping drift from the code.The same free-form shape, versioned by git, but nothing links a decision to the plan that took it.