SATE-Laravel
← All pipelines

Governance & Learning

sate learning

Internal

Multi-repo learning loop that extracts patterns into the canonical knowledge base.

What it does

Walks one or more local repositories, extracts test patterns via the refiner, merges them into the engine's knowledge base with provenance tagging (per-app seenInApps tracking, project labels, sanitized snippets). Pattern types include stable assertions, fragile patterns, mock setups, and factory chains.

Who runs this

  • →Internal SATE engineering (not part of the customer-facing surface today)

How to invoke

node refiner/localLearner.js --all-repos --repos-file refiner/localRepos.local.json

Flags

--all-repos

Iterate all repos defined in --repos-file.

--repos-file <path>default: refiner/localRepos.local.json

Local repo registry.

--testRoot <path>default: tests

Per-repo override.

What it produces

  • ·learning/refiner/canonical-kb.json — knowledge base with extracted patterns

What it does NOT do

  • ✖Does NOT run automatically — internal command for now
  • ✖Does NOT learn from non-approved tests (gated by MANUAL stage)
  • ✖Does NOT promote a pattern without sanitization (no customer literals leaked into snippets)

Notes

Surfaced here for transparency. Not part of the public sealed CLI today.

Status

Engine tests

0

Status

Internal

Category

Governance & Learning

Related pipelines

sate emit-learning-candidates

Opt-in: extracts STABLE_TESTHASH_PASS candidates from an execution trace for downstream review.

sate review-queue

Shows the REVIEW-class learning candidates a human needs to approve, reject, or defer.

sate submit-decision

Append a human decision (APPROVE / REJECT / DEFER) for a candidate to the append-only review ledger.