How selmakcby/a-sirketi works
A Şirketi is a Claude Code agent-workflow scaffold: Telegram links or scheduled work enter team queues, agents produce verification or analysis artifacts, a separate watchdog reviews runs, and the distributor may hand accepted work to another team. The main chain turns verified X material into a publication-ready X Article package; a person makes the final publish decision. YouTube analysis is a separate Apify-backed reporting lane. The README documents these product capabilities; sampled source excerpts confirm several orchestration links, but unsampled team prompts and output-generation internals are represented at their existing directory paths.
Main entry points: kos.py Team runner · takimlar Team queues and state · dagitici.py Chain distributor · telegram_dinle.py Telegram listener
Components 23
Intake and scheduling
- Telegram reader
bin/telegram_oku.py - Telegram listener
bin/telegram_dinle.py - Daily scheduler
bin/zamanla.py - Daily reports
bin/gunluk.py
Agent runtime and oversight
- Team runner
bin/kos.py - Agent definitions
bin/agents_uret.py - Run watchdog
bin/bekci.py - Chain distributor
bin/dagitici.py - NIM model proxy
bin/model_proxy.py
Content workflows
- X verification team
takimlar/x-icerik - X source fetcher
bin/tweet_cek.py - YouTube analysis team
takimlar/youtube-analiz - YouTube data collector
bin/youtube_analiz_cek.py - Apify actor client
bin/apify_cek.py - Weekly source data
- X Article team
takimlar/twitter-icerik - Cover generator
bin/kapak_uret.py - Publication package
takimlar/twitter-icerik
Queue and state
- Settings and limits
bin/ayar.py - Team queues and state
takimlar
Other components
- Operator
- Telegram bot
- Human publisher
Connections 30
- Operator → Telegram bot: sends link
README.md - Telegram listener → Telegram reader: fetches updates
bin/telegram_dinle.py - Telegram reader → Team queues and state: writes inbox
README.md - Telegram listener → Team queues and state: routes message
bin/telegram_dinle.py - Telegram listener → Team runner: starts team
bin/telegram_dinle.py - Telegram listener → Chain distributor: invokes after run
bin/telegram_dinle.py - Daily scheduler → Daily reports: schedules
bin/zamanla.py - Daily reports → Chain distributor: runs morning chain
bin/gunluk.py - Daily reports → Team queues and state: queues weekly work
bin/gunluk.py - Settings and limits → Team queues and state: reads and updates
bin/ayar.py - Team runner → Agent definitions: generates agents
bin/kos.py - Team runner → Run watchdog: calls fallback review
bin/kos.py - Team runner → Team queues and state: records run state
README.md - Team runner → X verification team: runs team
README.md - Team runner → YouTube analysis team: runs team
README.md - Team runner → X Article team: runs team
README.md - Run watchdog → Team queues and state: records decision
README.md - Chain distributor → Team queues and state: reads and writes queue
bin/dagitici.py - Chain distributor → Team runner: starts eligible team
README.md - X verification team → X source fetcher: fetches source
README.md - X verification team → Chain distributor: queues accepted item
README.md - Chain distributor → X Article team: dispatches accepted item
README.md - YouTube analysis team → YouTube data collector: requests analysis data
README.md - YouTube data collector → Apify actor client: runs actors
bin/youtube_analiz_cek.py - YouTube data collector → Weekly source data: writes weekly data
README.md - YouTube analysis team → Weekly source data: reads source data
README.md - X Article team → Cover generator: requests cover
README.md - X Article team → Publication package: produces package
README.md - Publication package → Human publisher: awaits publication
README.md - Team runner → NIM model proxy: uses NIM route
README.md