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
2026-09-23 12:33:58 +02:00
2026-09-25 21:41:17 +02:00

roledesk

A job-search engine for one person. It reads the public job boards of the companies you follow (Greenhouse, Lever, Ashby, Workable, SmartRecruiters, Recruitee, Personio) and runs saved searches on boards that span companies (Jobindex, Adzuna), keeps every posting with when it was first seen and when it closed, scores the open roles for each of your search agents, and tracks the applications you make.

Built on the boilerplate: its job queue and scheduler (pipeline/), password (auth/) and deploy kit (deploy/), plus the plugin contract and tools from news (desk/). FastAPI, Jinja2, SQLite, uv, no build step.

Quick start

uv sync --all-extras
uv run app companies seed      # 186 companies from seed/companies.json, one board each
uv run app job pull            # read every board and search, embed and score them, extract the best, rescore, summarise
uv run app serve               # http://127.0.0.1:8790

uv run app worker drains the queue without the web app, and uv run pytest runs the app's tests and the copied modules' own. config.yaml is the committed default, with one starter agent and four Copenhagen searches on Jobindex; config.example.yaml documents every key, including the summarise: block below.

Sources: job boards and company boards

The Sources sub-tab does all of this; the same actions exist on the command line:

uv run app searches add jobindex "Copenhagen golang" --q "golang go-udvikler" --area 1
uv run app searches add adzuna "Berlin backend" --q backend --country de --where Berlin
uv run app searches list
uv run app companies prune --in DK,SE,NO,EU      # lists; add --apply to unsubscribe

A Jobindex query is its own syntax: words are OR, "a b" is a phrase, +word is required; --area 1 is Region Hovedstaden. An Adzuna search needs a free key from developer.adzuna.com in ADZUNA_APP_ID and ADZUNA_APP_KEY. searches add runs the search once and saves it only if it finds something. Switching a board or search off closes its open roles; the next pull after switching it back on reopens the ones it still lists. The embeddings extra is what fit needs, agent is for extract on a model, passkeys for the Account tab's passkeys.

The tabs

Two tabs, and a cog for settings.

Roles newsdesk's feed: a filter rail (Show read, Best first or Newest first, first seen, tags, more filters), the open roles with company, place, age, tags and a summary, votes (▲ like ● neither way ▼ pass) and a Save button. A title opens the reader beside the list; on a phone it takes the screen. j/k move, Enter opens, / searches
Applications the saved roles and their status, each change recorded as an event

The cog opens /settings: a rail of pages on the left, the page beside it, sticky so Runs is one click away however long the company list gets. On a phone the rail is a strip under the header.

Search agents a card per agent, newest first: name, an On/Off switch, what it looks for, and its weights. Edit turns the card into the form in place (name, intent, weights, require/prefer gates behind an Advanced disclosure); New agent puts a blank card on top. Saving or switching on rescores; deleting one takes its votes, liked companies and scores with it
Sources one page showing one section at a time, picked in the rail or by the URL fragment: job boards (queries on Jobindex or Adzuna: try, save, edit, remove), company boards (every company whose board is read, its open roles and its affinity for the picked agent, following a company, pruning boards by region) and runs (Pull now, what is running, the schedule, one stage at a time, and the message log)
Account the password, sessions and passkeys. Open on loopback, locked on anything else

Likes, hidden companies and votes belong to a search agent, not to the account: two agents rank the same roles differently, and one may hide a company the other favours. An agent also learns from votes: its taste weight scores a role by how close it is to the roles the agent liked minus the ones it passed.

The reader, /roles/{uid}, shows the full posting: the same vote and Save controls, the company's own affinity, the tags and the summary when there is one. Opening a role marks it read, and so does a vote. A search role's posting is a teaser; the reader offers a Fetch button that scrapes the original listing once and stores the text. A Summarise button asks for a short summary of that one role; config.yaml's summarise: block turns summarising on for the pull itself, bounded to the top per_agent roles of each agent so the cost stays flat.

Docs

S
Description
No description provided
Readme
4.6 MiB
0 Stars 1 Watchers 0 Forks
Languages
Python 85.3%
JavaScript 10.4%
CSS 2%
HTML 1.5%
PowerShell 0.5%
Other 0.3%