Read role pages in an enrich stage after scoring, best roles first
The pull reads listings only. `enrich` (app/jobs/enrich.py) takes the open roles without text that an enabled agent ranks ungated in its best `enrich.top` at `enrich.min_score` or above, best score first, on boards whose reader has `detail`, within `enrich.per_board` and `enrich.per_run`, one batch per board. `plugins.ats.read_board_details` reads them and keeps the per-board state; a read that gave no text is read again after `enrich.retry_days` (finding 18). Filled roles are upserted, which drops their vectors, and embed and score follow; `cascade.step` passes over both when nothing was filled. The pull cascade is pull, embed, score, enrich, embed, score, extract, score, summarise. Saving an agent and changing a company's affinity queue the `enrich` job (enrich, embed, score) behind their rescore; a vote does not. The stage has a switch, `stage:enrich`, over `enrich.enabled`. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
3adafd8950
commit
77b45a4b0f
10 files changed
+956
-219
No files matched your search
+11
-9
@@ -1053,7 +1053,9 @@ def test_the_pull_job_runs_pull_embed_score_extract_score_then_summarise(
|
||||
assert batch is not None
|
||||
jobs = sorted(queue.jobs.of_cascade(batch.cascade_id), key=lambda j: j.id)
|
||||
stages = [j.result.get("task") for j in jobs if j.kind == FINAL]
|
||||
assert stages == list(cascade.CASCADES["pull"])
|
||||
# Enrich found no board with role pages, so the embed and score behind it
|
||||
# are passed over.
|
||||
assert stages == ["pull", "embed", "score", "enrich", "extract", "score", "summarise"]
|
||||
work = [j.kind for j in jobs if j.kind != FINAL]
|
||||
# The `pull` job and the chain it asked for are one cascade, so the Jobs
|
||||
# tab shows a scheduled pull as one run. Extract and summarise are off and
|
||||
@@ -1107,10 +1109,10 @@ def test_the_model_batches_are_isolated_with_a_timeout() -> None:
|
||||
def test_follows_walks_the_cascade_by_position_and_stops_at_its_end() -> None:
|
||||
"""By position, not by name: `score` is in the chain twice, once before
|
||||
extract and once after, and the second must lead on to summarise."""
|
||||
assert cascade.CASCADES["pull"] == ("pull", "embed", "score", "extract", "score",
|
||||
"summarise")
|
||||
assert [cascade.follows("pull", at) for at in range(6)] == \
|
||||
["embed", "score", "extract", "score", "summarise", ""]
|
||||
assert cascade.CASCADES["pull"] == ("pull", "embed", "score", "enrich", "embed", "score",
|
||||
"extract", "score", "summarise")
|
||||
assert [cascade.follows("pull", at) for at in range(9)] == \
|
||||
["embed", "score", "enrich", "embed", "score", "extract", "score", "summarise", ""]
|
||||
assert cascade.follows("pull", -1) == "" and cascade.follows("pull", 9) == ""
|
||||
assert cascade.follows("nope", 0) == ""
|
||||
assert cascade.chain("pull").carries == {"cascade": "pull", "task": "pull", "at": 0}
|
||||
@@ -1124,10 +1126,10 @@ def test_step_hands_the_position_on(monkeypatch: pytest.MonkeyPatch) -> None:
|
||||
made.append(t)
|
||||
return t
|
||||
|
||||
out = cascade.step(None, None, Tally(), {"cascade": "pull", "task": "score", "at": 2},
|
||||
out = cascade.step(None, None, Tally(), {"cascade": "pull", "task": "score", "at": 5},
|
||||
context=Context())
|
||||
assert out == {"task": "score", "next": "extract"}
|
||||
assert made[0].carries == {"cascade": "pull", "task": "extract", "at": 3}
|
||||
out = cascade.step(None, None, Tally(), {"cascade": "pull", "task": "score", "at": 4},
|
||||
assert made[0].carries == {"cascade": "pull", "task": "extract", "at": 6}
|
||||
out = cascade.step(None, None, Tally(), {"cascade": "pull", "task": "score", "at": 7},
|
||||
context=Context())
|
||||
assert out["next"] == "summarise" and made[1].carries["at"] == 5
|
||||
assert out["next"] == "summarise" and made[1].carries["at"] == 8
|
||||
Reference in new issue
Block a user