GoldenMatch

Add a Learning Memory correction. Two shapes: - pair-level: decision='approve' or 'reject', requires id_a + id_b - field-level (v1.18.2+): decision='field_correct', requires cluster_id + field_name + corrected_value Source is 'agent' with trust=0.5 (lower than human steward 1.0). Pair (id_a, id_b) is canonicalized to (min, max) before storage.

Approve or reject a review queue pair

Compare ER strategies on your data

Run full ER pipeline with confidence gating and reasoning

Explain why records are in the same cluster

Natural language explanation for a record pair

Match two files with intelligent strategy selection

Get borderline pairs awaiting approval

Diagnose blocking on the loaded dataset: returns ranked blocking key candidates with block counts, max block size, total candidate comparisons, and estimated recall. Use it to explain why matching is slow or produces too many candidate pairs.

Profile data, detect domain, recommend ER strategy

Run AutoConfigController on a CSV; return the committed GoldenMatchConfig (incl. negative_evidence / Path Y when chosen) plus telemetry — stop_reason, health, decision trace, indicator column priors. Programmatic equivalent of `goldenmatch autoconfig`.

Estimate match RECALL without ground truth (unsupervised). Treats each auto-configured matchkey/pass as a decorrelated system and uses capture-recapture over their overlaps to estimate how many true matches were missed. Returns a point estimate (a safe lower bound additionally needs a small labelled audit; see `goldenmatch evaluate --certify --audit-out`). Needs >=3 decorrelated systems.

Compare two ER clustering outcomes on the same dataset without ground truth (CCMS): classifies each cluster as unchanged / merged / partitioned / overlapping and returns the Talburt-Wang Index. Both inputs are JSON cluster files (as written by export-style output).

Diagnose weaknesses in the loaded run's auto-config: columns admitted that shouldn't be (source/provenance labels, per-row IDs), oversized or shared-value blocks, null sinks, low-signal matchkeys, and over-merging. Returns ranked findings, each with a plain-English explanation + a concrete fix, plus a one-paragraph summary.

Return the AutoConfigController telemetry from the most recent `auto_configure` or `agent_deduplicate` call in this MCP session. Same JSON shape as the web /api/v1/controller/telemetry endpoint.

Create a custom domain extraction rulebook. Define patterns for a specific data domain (medical devices, automotive parts, real estate, etc.).

Alias for `find_duplicates`. Find duplicate matches for a record. Provide field values to search against the loaded dataset.

Extract records from documents (PDF/image) against a target schema into rows ready for dedupe_df. Returns records + an ingest report.

Propose a target extraction schema (JSON) from a sample document image/PDF.

Score the loaded run against ground-truth pairs. Loads a ground-truth CSV (id_a,id_b columns) and returns precision, recall, and F1 for the current clustering.

Alias for `agent_explain_cluster`. Explain why records are in the same cluster

Explain why two records match or don't match. Shows per-field score breakdown.

Alias for `explain_match`. Explain why two records match or don't match. Shows per-field score breakdown.

Human-readable explanation of why each stage is routed the way it is, with the driver-RAM projection that drove it.

Export matching results to a file (CSV or JSON).

Find duplicate matches for a record. Provide field values to search against the loaded dataset.

Run GoldenCheck scan and apply fixes to a CSV file. Returns the fixed data summary and a manifest of all fixes applied. Requires goldencheck: pip install goldenmatch[quality]

Get details of a specific cluster: all member records and their field values.

Get the merged golden (canonical) record for a cluster.

Get dataset statistics: record count, cluster count, match rate, cluster sizes.

Export the append-only identity audit log in commit order: every event with actor / trust / timestamp / reason, so a reviewer can reconstruct exactly which actor changed what, when, and why. Optionally filtered by dataset / actor.

Anchor the append-only audit log with a tamper-evidence seal: a chained sha256 root over every event since the last seal. Cheap and idempotent (a no-op when nothing new has been logged). Run it periodically (or after a batch of stewardship actions) so the history becomes provably untampered. Optionally scoped to a dataset. Publish/mirror the returned root_hash to make tampering detectable by an external party.

Verify the append-only audit log against its seal chain. Replays the per-event content hashes and the seal roots to detect content edits, deletion, reordering, and insertion of any sealed event. Returns {ok, events_checked, seals_checked} plus the ids of any content mismatches / broken seals / missing sealed events. Optionally scoped to a dataset.

Claim a record into an identity, moving it out of any prior entity ('this record belongs to that identity'). Emits a provenance-stamped `claimed` event on both the gaining and losing entities.

List evidence edges marked `conflicts_with`.

Return the temporal event log for an identity.

List identities, optionally filtered by dataset/status.

Manually merge two identities. All records from `absorb_entity_id` are reassigned to `keep_entity_id`. The merge events are stamped with `actor`/`trust` provenance so the audit log records who merged these and on what authority.

MDM profile of one entity: record count + per-source breakdown, golden record, confidence, conflict count, canonical version (structural-event count), and first/last activity. Returns {found: false} when no such entity exists.

Resolve a record_id to its durable identity. Returns the full identity view (members, evidence edges, recent events) or null when no identity exists for that record.

Adjudicate a `conflicts_with` pair: 'same' keeps the entity intact, 'distinct' splits the second record out into a new identity, 'defer' only logs. Records a durable mediation verdict + event with actor/trust provenance, and stops the conflict re-surfacing in the open-conflicts queue.

Fetch the full detail of one identity by entity_id: its member records, evidence edges, and recent event log. Returns {found: false} when no such entity exists.

Split a subset of records off an identity into a brand-new identity. The original keeps the remaining records. The split events carry `actor`/`trust` provenance.

Graph-level summary / health stats: entities by status, total records, records-per-entity distribution, conflict total, source mix, and the largest entities. Optionally scoped to a dataset.

Prioritized steward worklist: active entities needing attention (open conflicts and/or confidence below weak_confidence), highest conflict count first.

Match a batch of new records against an existing base dataset (without re-running the whole base). Returns matched (new_row_id, base_row_id, score) pairs plus counts. Auto-configures from the base file if no config is given.

Force a MemoryLearner pass over accumulated corrections. Returns the list of LearnedAdjustments produced (matchkey_name, threshold, sample_size, learned_at). Requires >= 10 corrections per matchkey before threshold tuning fires; otherwise returns an empty list.

Field-level provenance for the loaded run: for each scored pair, the per-field scores that produced the match, plus cluster id. Optionally write a lineage JSON to a directory.

Flag config/env overrides that force a slow path (e.g. CLUSTERING_THRESHOLD=0 when the edge set fits driver RAM). ERROR at scale; would_refuse mirrors the runtime guard.

List duplicate clusters found in the dataset. Returns cluster IDs, sizes, and member counts.