Context & Session

Pre-computed graph analytics in one call: counts, health, chain completeness, business-area coverage, lifecycle balance. ~500 tokens vs ~5-8K for equivalent manual fetches.

Product summary, entity counts by type, and a human-readable graph overview. Call first to understand the file. Pass include_summary for edge counts, orphans, and edges-by-type.

Read session context: which skills ran, what was recommended, current focus area. Returns `recommendations_to_avoid`; the deduped list of recommendations already given this session. Pick your next recommendation NOT in that array (data-layer dedup, not prose).

Zero-state on-ramp: "there is nothing here yet, where do I begin?". Reads the live graph and, for an empty or barely-started graph, recommends the first canonical playbook (from UPG_PLAYBOOKS) plus the exact create_node call for its anchor entity. Established graphs are routed to plan / inspect / get_graph_digest instead. Takes no arguments.

Send feedback about the Unified Product Graph (a bug, a feature request, an observation) to the project's triage queue at unifiedproductgraph.org, from any MCP client. Anonymous; no account. SHAPE IT FIRST for actionability, asking the user at most one round of questions (don't guess): a bug wants steps to reproduce, expected vs actual, and severity (`details`); a feature_request wants the underlying problem, the desired outcome, and any current workaround (`details`). CONSENT IS REQUIRED: the tool sends NOTHING unless you pass confirmed:true. Call it first WITHOUT confirmed to get back the exact payload, show that payload to the user (including the auto-collected `context`), and only re-call with confirmed:true after they say yes. PRIVACY: `context` is auto-assembled from the MCP client name/version, the server version, the runtime, and graph SIZE counts only; it never includes node titles, descriptions, or any graph content.

Update session context: register a skill invocation, record a recommendation, set focus area, switch lens, or store custom state for cross-skill coordination.