Score roles per search agent, and run pull into extract, embed and score

embed stores one vector per open role; score rates every open role for
every enabled agent from require and prefer gates, fit against the
agent's intent, recency and company affinity, keeping gated rows with a
reason. Schema v3 adds vectors and the cached intent vector. The page
gets an agent order, the score on each row, and an Agents tab. Two
starter agents in roles.yaml are imported once.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
WolfandClaude Opus 5.5 committed 2026-09-23 12:14:21 +02:00
1 parent 4dfb8831aa
commit bc60324fab
22 files changed
+1457 -44

No files matched your search

+66 -1
View File
@@ -15,7 +15,7 @@ import re
import unicodedata
from typing import Annotated, Any, Literal, get_args
from pydantic import BeforeValidator
from pydantic import BeforeValidator, Field
from desk.models import JsonDict, Model, ModelTaskResult
@@ -152,6 +152,14 @@ class ExtractResult(ModelTaskResult):
return self.extracted
class ScoredRole(Role):
"""A role as `/api/roles` serves it: with one agent's score and gate
reason when an agent was asked for, and without otherwise."""
score: float | None = None
gated_reason: str = ""
class Company(Model):
key: str
name: str
@@ -202,6 +210,63 @@ class Application(Model):
updated_at: str = ""
# ------------------------------------------------------------------ agents
#: The role columns a gate can name. `company` matches on the company key;
#: `place` is a case-insensitive substring, like the place filter; the rest are
#: case-insensitive equality, except `salary_min` (at least the value) and
#: `languages` (every language the role requires is one of the values).
GateField = Literal["place", "country", "remote", "seniority", "function", "salary_min",
"currency", "permit", "languages", "company"]
GATE_FIELDS: tuple[str, ...] = get_args(GateField)
GateMode = Literal["require", "prefer"]
class Gate(Model):
"""One condition of an agent. `require` gates a role that fails it (the
row is kept with the reason); `prefer` adds `weight` to its score when it
matches."""
field: GateField
values: list[str | int | float] = Field(min_length=1)
mode: GateMode = "require"
weight: float = Field(1.0, ge=0)
class Weights(Model):
"""How much each score term counts. Summed with the prefer gates' weights
into the denominator, so a score stays in 0..1 whatever the numbers."""
fit: float = Field(0.6, ge=0)
recency: float = Field(0.2, ge=0)
affinity: float = Field(0.2, ge=0)
class Agent(Model):
"""A row of `agents`: a named search with its own intent, gates and
weights. The cached intent vector is a column this does not declare."""
id: int
name: str
intent: str = ""
gates: Annotated[list[Gate], BeforeValidator(lambda v: _json(v, "[]"))] = []
weights: Annotated[Weights, BeforeValidator(lambda v: _json(v, "{}"))] = Weights()
enabled: bool = True
created_at: str
class Candidate(Model):
"""What scoring reads about one open role for one agent: when it was
first seen, its company's affinity, and per gate `1` match, `0` no
match, `-1` the column is empty."""
uid: str
first_seen: str
affinity: int = 0
gates: list[int] = []
class ApplicationEvent(Model):
id: int
application_id: str