Fase 3, con una corrección sobre lo que decía la §11 de la spec de fase 2. El bloqueo no es OAuth. Los vídeos subidos por videos.insert desde un proyecto de API sin auditar quedan restringidos a privado, y el candado es del proyecto, no del vídeo: no se abre desde Studio, se abre pasando la auditoría de cumplimiento de Google. Así que esto no publica. Deja el vídeo en el canal con los metadatos puestos y devuelve el enlace de Studio para que una persona lo revise y le dé a publicar — la misma forma que /publish con los borradores de Ghost, y por la misma razón: el informe de fundamento no sirve de nada si el vídeo ya está subido cuando lo lees. Comando aparte, no un paso de /generate short_en. - src/generator/youtube.py: refresco de token contra oauth2.googleapis.com, subida resumable en dos pasos y metadatos derivados del shot spec ya guardado (título, enlace al artículo, fuentes que el Short cita en pantalla, etiquetas del tema). Sin google-api-python-client: es síncrono y bloquearía el loop del bot; son dos peticiones HTTP y el repo ya firma los JWT de Ghost a mano. aiohttp con SAFE_ACCEPT_ENCODING como todo lo demás. - Scope youtube.upload y nada más: un token filtrado no puede leer ni borrar nada del canal, sólo subir. - forced_private detecta que YouTube devolvió "private" cuando se pidió otra cosa, y el aviso lo dice. Es la firma del candado, y tragárselo haría creer que salió publicado. - invalid_grant se traduce a su causa real: la pantalla de consentimiento quedó en "Testing" y Google revoca esos tokens a los siete días. Es el fallo que menos se adivina y el que más probable es encontrarse. - get_article_url ahora excluye las filas short_en. Su published_url pasa a ser la URL de YouTube, y sin el filtro el siguiente Short de la sesión enlazaría al Short anterior: un bucle silencioso, porque la URL es válida y nadie la mira dos veces. - scripts/youtube_oauth.py, sólo stdlib: corre en el portátil, no en el contenedor, y no debería exigir instalar nada. - Las tres claves van optional:true en el Deployment. Sin eso, una clave que aún no está en Infisical deja el pod en CreateContainerConfigError y tira el bot entero por una función que nadie ha pedido todavía. 30 tests nuevos contra un servidor falso. No hay test en vivo a propósito: cualquier ejecución real sube un vídeo a un canal de verdad, y eso no es algo que deba pasar por teclear pytest. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
8.3 KiB
🦉 ResearchOwl
Exhaustive research engine with Telegram interface.
Recursively discovers, scrapes, and processes sources from across the web, then generates podcast scripts, blog posts, reports, or social threads using Ollama.
Architecture
Telegram (/research <topic>)
↓
ExhaustiveScraper
├── DuckDuckGo (8 queries × 5 results)
├── Wikipedia + recursive internal links
├── Reddit (top posts + top comments)
├── YouTube (transcripts)
├── PDFs (public documents)
└── Web scraping (trafilatura)
↓ recursive expansion (depth 1-3)
ContentProcessor (Ollama qwen2.5:7b + bge-m3 embeddings)
├── Chunking (800 token chunks, 100 overlap)
├── Quality scoring (0-10 per chunk)
├── Embeddings (cosine similarity RAG)
└── Deduplication
↓
OutputGenerator (Ollama)
├── 🎙️ Podcast script (20-30 min)
├── 📝 Blog post (1500-2500 words)
├── 📊 Research report (structured)
└── 🐦 Social thread (15-25 tweets)
Telegram Commands
| Command | Description |
|---|---|
/research <topic> |
Start exhaustive research |
/status |
Check progress |
/finish |
Stop early, proceed to generation |
/generate podcast|blog|report|thread |
Generate output |
/generate short_en |
Vertical Short: shot spec → grounding check → MP4 |
/short_spec |
Last shot spec as a JSON file, to hand-edit and re-render |
/upload_short |
Upload the rendered Short to YouTube (private, for review) |
/sources |
List all sources found |
/cancel |
Cancel current research |
Shorts (/generate short_en)
Claude writes a shot spec — typed JSON, not prose — which shortsmith renders into a 1080×1920 MP4. The bot sends the video and, in a separate message, a claims report.
/research JAL 1628 Alaska 1986 …
/generate blog en → Ghost draft, article URL stored on the output row
/generate short_en → spec → grounding → render → video + claims report
/upload_short → uploads to YouTube as PRIVATE, with metadata filled
in; publishing stays a human click in Studio
Three things make this different from generating text, and each has its own mitigation:
- It is a contract, not prose. The template schemas are fetched live from
GET /templatesand never copied here, so a template added to shortsmith is available immediately. A spec is validated locally against those schemas before anything renders, and the exact error paths (shots.0.radar_sweep.props.sweeeps) go back to the model verbatim — up to 3 attempts. - It contains figures and quotes.
grounding.pyextracts every quote, figure, date and proper noun and checks it against the exact chunks the model was given. No LLM in that path: normalisation plus substring, deterministic and free. Whatever is not in the chunks is checked against the worked example that travels in the prompt, so a figure lifted from it is reported as a prompt leak, not as an invention — different diagnosis, different fix. Neither ever blocks the render: both are surfaced next to the video and a human decides. - It becomes a published video.
/generate short_enuploads nothing: the MP4 lands in Telegram for review and in/data/shorts/{session_id}.mp4. Getting it onto the channel is a separate, explicit/upload_short.
Fallbacks hold throughout: if shortsmith is unreachable, the job errors, or the spec never validates, the spec JSON comes back as a file. The expensive part is the generation, not the render.
Full spec of the phase: docs/shortsmith-phase2-spec.md.
YouTube (/upload_short)
Uploads /data/shorts/{session_id}.mp4 to the channel with the title from the
spec, a description carrying the article link and the sources the Short cites on
screen, and tags derived from the topic. The YouTube URL is written back to the
output row, so a second /upload_short on the same session refuses unless you
say /upload_short force.
Read this before setting it up. Videos uploaded through videos.insert from
an unaudited API project are restricted to private viewing
mode. The lock
belongs to the API project, not to the video — you do not unlock it from Studio,
you unlock it by passing Google's compliance audit. So this command does not
publish. It puts the video on the channel with the metadata already filled in
and hands back the Studio link; a person reviews and presses publish. That is
the same shape as /publish, which only ever writes Ghost drafts.
One-time setup, in console.cloud.google.com:
- Enable YouTube Data API v3 on a project.
- OAuth consent screen → External → publish it to "In production". Leaving it in "Testing" makes Google revoke the refresh token after seven days, and the bot dies on its own the following Tuesday.
- Credentials → OAuth client ID → Desktop app.
python scripts/youtube_oauth.py --client-id … --client-secret …, which opens a browser, catches the redirect on localhost and prints the refresh token.- Put
youtube-client-id,youtube-client-secretandyoutube-refresh-tokeninto Infisical (they arrive asresearchowl-secrets-infisical).
The scope requested is youtube.upload only: a leaked token cannot read or
delete anything on the channel — the worst it can do is upload. Quota is not a
concern (1 unit per upload, 100 uploads a day). YOUTUBE_ENABLED=false is the
kill switch.
Local Development
# 1. Clone and setup
git clone https://git.chemavx.xyz/chemavx/researchowl
cd researchowl
# 2. Create virtualenv
python3 -m venv venv && source venv/bin/activate
pip install -r requirements.txt
# 3. Configure
cp .env.example .env
# Edit .env with your values
# 4. Run
python main.py
Deploy to k3s
# 1. Create namespace and secrets
kubectl create namespace researchowl
kubectl create secret generic researchowl-secrets \
--from-literal=telegram-bot-token=YOUR_TOKEN \
--from-literal=telegram-allowed-users=YOUR_USER_ID \
-n researchowl
# 2. Copy manifests to your k8s-manifests repo
cp k8s/*.yaml /path/to/k8s-manifests/researchowl/
# 3. Apply ArgoCD app
kubectl apply -f k8s/argocd-app.yaml
# 4. Push to Gitea → Gitea Actions builds → ArgoCD deploys
git add . && git commit -m "feat: add researchowl" && git push
Tuning
| Variable | Default | Description |
|---|---|---|
MAX_SOURCES |
150 | Hard cap on sources |
MAX_DEPTH |
3 | Link recursion depth |
QUALITY_THRESHOLD |
0.4 | Min chunk quality (0-1) |
REQUEST_DELAY |
1.0s | Delay between requests |
Want more thoroughness?
- Increase
MAX_SOURCESto 300+ - Increase
MAX_DEPTHto 4-5 - Lower
QUALITY_THRESHOLDto 0.3
Want faster results?
- Lower
MAX_SOURCESto 50 - Set
MAX_DEPTHto 1-2 - Higher
QUALITY_THRESHOLDto 0.6
Bot avatar
The profile picture of @chemavx_researchowl_bot is not an opaque binary
checked into the repo: assets/make_avatar.py draws it with PIL at 4× and
scales it down, so the emblem can be retouched without hunting for an original.
Telegram crops avatars to a circle, so everything that matters lives inside
the inscribed circle; verified legible at 48 px.
python3 assets/make_avatar.py # writes assets/avatar.png
It is applied over the API with the token from the secret, no BotFather.
Watch out for setMyProfilePhoto: its photo parameter is not the file, it is
an InputProfilePhoto object pointing at the attachment. Posting the file on
its own gets you a baffling photo isn't specified.
TOK=$(kubectl get secret researchowl-secrets-infisical -n researchowl \
-o jsonpath='{.data.telegram-bot-token}' | base64 -d)
curl -s -F 'photo={"type":"static","photo":"attach://av"}' \
-F "av=@assets/avatar.png" \
"https://api.telegram.org/bot$TOK/setMyProfilePhoto"
unset TOK
Notes
- Uses qwen2.5:7b (scoring) and bge-m3 (embeddings) on your existing Ollama — zero API cost
- Optionally add
ANTHROPIC_API_KEYfor Claude fallback on generation - SQLite database stored in
/data/researchowl.db - All outputs saved to DB and available via
/outputs