Component · spec & evidence

Reference Knowledge Routing

Runs a keyword-overlap match that ranks catalog entries, shows why each matched, and admits when none do.

Contract

Does Give it a plain-English problem and a small catalog of reference entries (each tagged with the domains, capabilities, and problem-spaces it covers), and it returns the entries most likely to help, ranked, with a breakdown of why each matched. It weighs strong structured tags more than loose description text, and shows its working rather than hiding it behind a black-box score. If nothing in the catalog actually overlaps your problem, it says no_match instead of inventing a result. It is a transparent keyword-overlap matcher, not a search engine or embedding model, and it never reaches outside the catalog you hand it.

Scope limit The bundle demonstrates explainable tiered weighted-token retrieval over a sanitized reference catalog. It is not BM25, TF-IDF, embedding search, repository cloning, license authority, private reference corpus authority, or launch-scope decision.

Run

microcosm annex-knowledge-routing run --input fixtures/first_wave/annex_knowledge_routing/input --out /tmp/plectis-public-runs/annex_knowledge_routing

microcosm annex-knowledge-routing run --input fixtures/first_wave/annex_knowledge_routing/input --out receipts/first_wave/annex_knowledge_routing --acceptance-out receipts/acceptance/first_wave/annex_knowledge_routing_fixture_acceptance.json

Evidence

ClassBounded runtime computationevidence 4/5Real runtime resultRuns real tools

Links to Certificate Kernel Execution Lab, Verifier Lab Kernel, Formal Math Lean Proof Witness

navigationarchitecture

Source

Source Design note · Source

Map Show on the map

Index Open compact index card

Paper module