Commit Graph
28 Commits
Author SHA1 Message Date
WolfandClaude Opus 5.5 6d82287402 Bring the Python side to the ADR: the judge scores 0 to 10 with the cap at 4 and migration 33 clears every judgement, summarise reads the top 30 of the want and the match order, a hidden company's roles are never in the feed, and the company switch is called All locations.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-04 21:40:32 +02:00
WolfandClaude Opus 5.5 8b6c6d5ac3 Test migration 32, the cut, saved filters and the override rescore, and bring the READMEs and config comments to fetch text and the cut
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-04 15:16:38 +02:00
WolfandClaude Opus 5.5 617356e60c Rank by the cut: fetch gate overrides in scoring, the cut per agent, stages read the cut, fetch text renamed from enrich, the feed kept to the fetch gate, and migration 32
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-04 15:11:43 +02:00
WolfandClaude Opus 5.5 9960799b96 Judge each agent's best roles against its CV as skills, experience and career, a stage after summarise
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-03 22:00:18 +02:00
WolfandClaude Opus 5.5 8c6ea0ffb6 Leave out titles by phrase: an exclude gate mode and a title gate field
A role whose title holds one of the agent's phrases as whole words is gated
with the phrase as its reason, a favourite's too. The agent form has a
Leave out titles line under Where. function = sales could not do it, since
the classifier files Solutions Engineer and pre-sales under sales.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-30 15:27:43 +02:00
WolfandClaude Opus 5.5 d04ceae7e0 Prep stage: host words stored raw, mapped centrally in app/prep after every pull
upsert_role stores the host's own words (roles.*_host), the place texts
unresolved when they change, and the company key unresolved on insert, and
marks the role prep_due. app.prep maps seniority, employment, remote and
country through the host tables in app/prep/hosts, resolves places and
company aliases, lists unknown words in `unmapped`, remaps every open role
when the tables' fingerprint changes, and runs rebuild_dupes, which moves
out of the pull's settle. Migration 22 adds the columns and the table and
fills the host words from raw. Cascade: pull, prep, embed, score, enrich,
prep, embed, score, extract, score, summarise. `app prep [unmapped]
[--remap]`, /api/prep/unmapped and a fold on the Sources page.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-30 11:19:22 +02:00
WolfandClaude Opus 5.5 77b45a4b0f 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>
2026-09-30 09:53:29 +02:00
WolfandClaude 3ab8035d0d Gate a favourite by the require gates and drop its 0.9 floor
A favourite company skips the any group only; a require gate it fails
gates it with the usual reason. Its score is the weighted sum with the
affinity term at 1, and weights that sum to 0 give "no weights" as for
every other company.

Co-Authored-By: Claude <noreply@anthropic.com>
2026-09-29 21:52:26 +02:00
WolfandClaude Opus 5.5 2534803df3 Add app agents list and prompt, and document prompts for the seed agents
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 16:20:23 +02:00
WolfandClaude Opus 5.5 4e3c3f130e Give a search agent several prompts and score a role by its best one
Migration 14 moves each non-empty agents.intent into agent_prompts as one
prompt named after the agent, with its cached vector, and drops the three
intent columns. A role's fit is the maximum over the agent's prompts of the
calibrated cosine, and scores.prompt names the prompt that won. The agents
API takes prompts and still takes intent; the role endpoint serves the
agent's score and prompt.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-29 16:19:57 +02:00
WolfandClaude Opus b021838959 Calibrate fit and taste per embedding model instead of min-max per run
scale() maps a cosine to clip((cos - lo) / (hi - lo), 0, 1) with constants
from CALIBRATION in app/config.py, keyed by model and replaceable under
embed.calibration. multilingual-e5-small: fit 0.75..0.88, taste -0.14..0.19
(taste from synthetic votes). An uncalibrated model falls back to min-max
and logs a warning once per run.

The extract and summarise bars default to 0.55 and 0.65, about 300 and 35
roles for the starter agent on the measured data; stored limits are kept.
An agent without an intent is left out of the model-stage scope, and each
score run names such agents in one score.no_intent notice.

Co-Authored-By: Claude Opus <noreply@anthropic.com>
2026-09-29 14:59:34 +02:00
Wolf 43773efc53 Merge branch 'worktree-agent-a3bd292551395dda0'
# Conflicts:
#	app/storage/schema.py
2026-09-29 14:16:35 +02:00
Wolf c6246af994 Merge branch 'worktree-agent-ac3c9345edd29a1c6'
# Conflicts:
#	app/cli.py
#	app/jobs/sources.py
#	app/static/app.css
#	app/static/companies.js
#	app/web/api.py
#	tests/test_web.py
2026-09-29 14:15:10 +02:00
WolfandClaude Sonnet 75eca1c914 Date a role by posted_at, not just first_seen
recency and the newest feed both counted from first_seen, when the app
itself first stored a role, so a newly added board's whole backlog scored
as freshly posted for two weeks: 10,223 of 15,415 board roles were posted
more than 30 days before they were first seen (average gap in days: Ashby
78, Greenhouse 143, Lever 226, Personio 333).

Add roles.posted_at (migration 9): the earlier of first_seen and a trusted
host's published_at, where TRUSTED_POSTED_HOSTS (app/models.py) records
which of the ats and search hosts report a creation date rather than one
that moves on an edit, checked against live boards and the live database.
A future or pre-2015 date is not trusted either. upsert_role sets it on
insert and keeps the earliest on update, so an edit on the board can only
move it earlier, never later. recency, the feed's newest order, the "new
since" days filter and _top_scope's tie-break now read posted_at;
first_seen is unchanged and still what the UI shows as when the app found
a role. The row's age in roles.js now shows the posted date instead.

A partial index (posted_at DESC, uid) WHERE closed_at IS NULL serves the
newest feed; sqlite3's bundled build has no STAT4, so its planner
otherwise sorts every open role instead of using it, measured 30x slower
on the live database, so Store.roles() forces it for that query shape.

Co-Authored-By: Claude Sonnet <noreply@anthropic.com>
2026-09-29 14:13:06 +02:00
WolfandClaude Opus b43fc54565 One pull at a time: the run is claimed in the database and its marker rides in each batch
A second pull (schedule, Run button, `app job pull` in another process) could
start while one was fetching and overwrite the stored run marker, so the first
run's boards failed their stamp check or a settle closed on the other run's id
lists. `claim_run` sets `pull:run` in one conditional statement, only while it
equals `pull:settled` or is older than an hour; a refused `pull` job queues no
stage and says so in its result and a notice, and the CLI exits 1. Batches
carry the marker in their payload, the plugins stamp with it, and the settle
reads it off its own batches' payloads before releasing the run.

Co-Authored-By: Claude Opus <noreply@anthropic.com>
2026-09-29 14:12:17 +02:00
WolfandClaude Sonnet b560cdce2a Switch the embed model to multilingual-e5-small for the European corpus
Measured on the live roles against the "Copenhagen engineering" agent's
intent: the English-only all-MiniLM-L6-v2 separates engineering from
non-engineering roles with an AUC of 0.68 overall but only 0.74 restricted
to Danish-country roles (a Danish engineering role beats the rest's
non-engineering median only 68% of the time). multilingual-e5-small reaches
0.70 overall and 1.00 on Danish-country roles, for about 3x the CPU time.
paraphrase-multilingual-MiniLM-L12-v2 measured worse than the English-only
baseline (AUC 0.60) and was dropped.

e5 needs "query: "/"passage: " prefixes; app/jobs/embed.py adds them by
matching the model name so a model change can't leave a stale prefix
setting behind. Added tests for the prefixing and for the two invariants
the model change relies on: embed drops the old model's vectors before
re-embedding, and an agent's cached intent vector is keyed by model so a
model change is a cache miss, never a stale-dimension read.

Co-Authored-By: Claude Sonnet <noreply@anthropic.com>
2026-09-29 13:59:47 +02:00
Wolf 074a41c6cf Merge branch 'main' into isolate-batches 2026-09-28 19:42:51 +02:00
WolfandClaude Fable 5.1 b73a0fc2cb The model batches are isolated with a deadline, so a Kill or a hang ends them
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-09-28 19:42:28 +02:00
WolfandClaude Fable 5.1 fb83ae5099 Gates reject on evidence only, any gates are one group, remote is read off the location line
A require gate on an empty column gated the role ("country unknown"), and extract, the stage that fills the column, is bounded to the roles that rank, so 2,281 of 15,556 open roles could never come back: every board that writes "Remote", "Hybrid" or "Distributed" as its location. An empty column now neither passes nor fails a gate; the role ranks, extract reads it, the rescore decides.

`any` is a third gate mode: the any gates are one group a role must match one of, so a ranking can say "in Denmark, or remote", which require alone could not. The group is gated only when some member had evidence against it.

`remote_of(place)` fills `remote` from the location line the way `country_of` fills `country`, and the startup backfill does both, since Greenhouse writes "Hybrid" as the location and gives no flag.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-09-28 19:30:06 +02:00
WolfandClaude Fable 5.1 7db77e83a3 Extract is bounded to the ranking: pull → embed → score → extract → score → summarise
The free stages run first and their score bounds the two model stages.
Extract reads only the best extract.per_agent roles of each agent at
extract.min_score or more, then score runs again so the gates read what it
filled; summarise gets a min_score of its own. The cascade steps by
position, since score is in it twice.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-09-28 15:26:58 +02:00
WolfandClaude Fable 5.1 a0772519b3 Per-agent ranking: votes, affinity and a taste term; settings are three sub-tabs
Votes on roles and company affinity move to role_votes and agent_companies,
one row per search agent, so two agents rank the same roles differently. The
score gains taste, the cosine to mean(liked) - mean(passed) of the agent's
votes, as newsdesk ranks articles. The API takes agent on roles, votes and
companies and falls back to the default agent, the enabled one with the
lowest id.

Settings become Search agents (a list again, with New and Delete), Sources
(job boards, company boards and runs with the message log, one section per
plugin) and Account. Old paths answer 301 to their section.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-09-28 12:53:41 +02:00
WolfandClaude Fable 5.1 9035529e0d Summarise and fetch text: a summary stage after score, two role routes
app/jobs/summarise.py is a ModelTask beside extract: about 60 words per role
from title, company, place and the fetched page else the description (capped
at 8000 characters). The pull cascade gains a summarise stage after score that
takes only roles without a summary in the top summarise.per_agent of an
enabled agent's ungated scores, and finds nothing when summarise.enabled is
false or no model is reachable. A reply that leaves a role unanswered is
counted in the state table and the role is given up after three.

POST /api/roles/{uid}/text scrapes original_url once and stores
article_text; POST /api/roles/{uid}/summarise queues a summarise_batch for
that uid alone, deduped per role.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-09-25 21:39:43 +02:00
WolfandClaude Opus 5.5 7d31cca394 Two tabs and a cog: settings shell, one ranking, a script per page, the phone frame
The top bar is Roles and Applications plus health's cog. Behind the cog,
/settings/{ranking,searches,companies,pulls,messages,account} share one row
of sub-tabs, an underlined row on a desktop and pills on a phone. The old
paths answer 301 with their query string.

Ranking edits the first enabled agent, or creates one, with the intent,
the weights and the gates under Advanced. The order menu on Roles reads
Newest and Best match with one enabled agent, and names them with several.
config.yaml keeps the Copenhagen starter; the other two are commented
examples in config.example.yaml.

roles.js is split into roles.js, applications.js, ranking.js and
companies.js; count, cell, debounce, the company list and the affinity
control live once in app.js. app.css gets the phone frame after health and
newsdesk: one-row bar, edge-to-edge panels, 44px targets and 16px fields on
a coarse pointer, hover only where there is one, safe areas. The app is
installable from a manifest and roledesk-coloured icons. Every lede touched
is written for the person using the app.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-25 21:29:06 +02:00
WolfandClaude Opus 5.5 047f7cd8ad Find roles by search, not only by company: Jobindex and Adzuna searches
The ats boards only read companies someone listed, and the seed list is
US-centred: 31 of 29,000 open roles were in Denmark. A `search` plugin runs
saved queries on boards that span companies, pulled in the same task as the
boards: Jobindex (Denmark, RSS, no key) and Adzuna (UK and most of the EU,
free key, built to its documented API). Jobnet and Jooble were checked and
left out.

`country` is now filled from the board's location line by a lookup, with EU
for Europe-wide roles, so country gates work without extract. `app companies
prune --in` unsubscribes boards with no open role in a region. The starter
agents and searches are Copenhagen and Europe.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-23 19:55:50 +02:00
WolfandClaude Opus 5.5 8b246c3035 Scale fit across each agent's run, and lower the starter prefer weights
With fit as (cos + 1) / 2 the real run spanned about 0.55 to 0.79, so a
prefer gate at weight 0.2 moved a score more than the whole spread of fit.
fit is now the cosine min-max scaled across the roles an agent scores in
that run: the closest is 1, the farthest 0, a single role 1, and a role a
require gate rules out stays out of the range. The starter agents' prefer
weights drop to 0.05, enough to break a tie. A regression test pins the
case: two close cosines that the old mapping let a 0.05 gate reorder.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-23 12:51:08 +02:00
WolfandClaude Opus 5.5 59a301c351 Wire the app like the boilerplate: door, queue, schedules, eight tabs
app/main.py assembles it in the boilerplate's order, auth.connect first,
then the queue with the scheduler and the pages. TABS is Roles,
Applications, Agents, Companies, Jobs, Schedules, Messages and Account:
the alpha page's three views become templates on the boilerplate's
base.html, driven by static/roles.js (the old app.js) through api() and
toasts.js, over the unchanged JSON routes. Companies lists every board with
its open roles and the affinity control. A 422 is one sentence, so a toast
never reads [object Object].

Config takes the boilerplate's shape: APP_CONFIG or config.yaml, a missing
file is the defaults, an unknown key is an error. roles.yaml is gone;
config.yaml is the committed default and carries the starter agents.

pull is now a job that asks for the pull cascade, because a schedule
enqueues a job kind; its default schedule, every 6h, is declared beside it
in app/jobs/cascade.py. pull_batch's targets are keyword-only so the Jobs
tab offers no batch button. `app worker` is the boilerplate's; `app job`
still drains a queue of its own and fires no schedules.

test_coupling holds pipeline, auth and desk apart (desk may import
pipeline) and plugins to desk; test_architecture keeps app's layers, now
with main.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-23 12:48:12 +02:00
WolfandClaude Opus 5.5 7a63311714 Rename the package roledesk to app
The boilerplate's rule: the deploy kit's Dockerfile, the APP_CONFIG,
APP_PASSWORD and APP_DATA variables and its commands all assume a package
called app, so a project renames its display title and nothing else. The
console script is `app`, and the launch entry runs `uv run app serve --port
8790`. The domain noun stays `role`.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-09-23 12:30:10 +02:00
WolfandClaude Opus 5.5 bc60324fab 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>
2026-09-23 12:14:21 +02:00