Bring the ADR to the owner's decisions: the affinity bonus as the code computes it, judge on 0 to 10 with a cap at 4, and a role's own vote counting directly

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
WolfandClaude Opus 5.5 committed 2026-10-04 21:09:56 +02:00
1 parent a0cb8c6a71
commit 6cad9713ac
2 files changed
+10 -3

No files matched your search

+9 -2
View File
@@ -94,7 +94,7 @@ Each role is embedded and scored semantically:
- **Prompt closeness** is its level against the best of the agent's prompts.
- **CV closeness** is its level against the closest part of the CV.
Both are calibrated levels from 0 to 1. A liked company adds 0.05 to both levels, a favourite 0.1. The bonus counts before the count and the minimum apply.
Both are calibrated levels from 0 to 1. A company's affinity adds to both levels the share it has in the `want feed score`: the company like weight times the company like's rise above neutral (5.1.1), over the sum of the want feed score weights. On the start weights a liked company adds about 0.056, a favourite about 0.111. The bonus counts before the count and the minimum apply.
#### 2.2 Input analytics filter
@@ -140,7 +140,13 @@ Each role is extracted once. A judgement is redone only when its input changes:
#### 4.2.1 CV match
Judge rates how closely the job spec matches the experience in the CV on three dimensions, skills, experience and career, each from 0 to 5. The **CV match** is the three numbers weighted as given in the agent settings (30/25/30 to start). A must-have the role breaks caps career at 2. Roles with the same CV match are ordered by CV closeness.
Judge rates how closely the job spec matches the experience in the CV on three dimensions:
- skills,
- experience,
- career,
each from 0 to 10. A model's rating is an estimate, not a measurement, so a coarse scale serves better than 0 to 100. The **CV match** is the three numbers weighted as given in the agent settings (30/25/30 to start). A must-have the role breaks caps career at 4. Roles with the same CV match are ordered by CV closeness.
## 5 Feed
@@ -166,6 +172,7 @@ The user likes or passes on roles in the feed, per agent. Each role has an embed
- With only likes, the direction is the mean of the liked roles; with only passes, away from the mean of the passed roles.
- With no votes, `job like` drops out and its weight leaves the sum.
- A role's own vote also counts directly when the feed is shown: a liked role's want feed score is multiplied by 1.2 (at most 1), a passed role's by 0.5. A vote never hides a role.
### 5.2 UI
+1 -1
View File
@@ -1,4 +1,4 @@
"""The Roles page's part of the ranking flow (docs/architecture.md, "5 Feed"):
"""The Roles page's part of the ranking flow (documentation/ADR Fetch prep and score.md, "5 Feed"):
the note that the feed keeps to the agent's fetch gate, the saved filters in
the rail and the routes they call, as `static/roles.js` uses them. A temp
database and FastAPI's TestClient; no browser."""