dossier-briefing

name	dossier-briefing
version	1.0.0
category	assist
description	Generates a structured research briefing for a topic or person as a Markdown scaffold (stdout or file). No persistent store.
tags	["briefing","dossier","recherche","markdown","research"]
standalone	true
anthropic_compatible	true
bach_compatible	false
bach_origin	true
languages	["de","en"]
dependencies	{"python":["datetime","pathlib","textwrap"]}
runtime	python3
entry_point	dossier_briefing_core.py
provenance	{"origin":"BACH persoenlicher-assistent","origin_path":"system/agents/persoenlicher-assistent/tools/dossier_generator.py","origin_version":"1.0.0","origin_repo":"github.com/ellmos-ai/bach","origin_license":"MIT","last_sync_from_origin":"2026-06-22","last_sync_to_origin":"None","local_changes_since_sync":"Alle Origin-DB-Abhaengigkeiten entfernt (create_dossier, update_dossier, DOSSIERS_DIR, DossierGenerator-Klasse mit DB-Methoden). Nur _create_markdown-Logik portiert und verallgemeinert (Person→Subjekt). Kein Store. One-Shot-Scaffold-Generator. Headless, nur Stdlib.\\n"}
language	de

Deutsch — Offizielle Deutsch-Version / Documento Oficial en Deutsch.

Dossier-Briefing (Deutsch)

Structured research briefing for a topic or person

---

Übersicht & Zweck

Generates an empty, structured Markdown briefing for any subject (person, company, event, concept). The scaffold serves as a starting point for subsequent research with research-agent or web-reading.

---

Triggers

Phrase	Action
"Create a briefing on Marie Curie"	Scaffold: person, type=person
"Dossier on OpenAI"	Scaffold: company, type=organization
"Briefing on quantum computing"	Scaffold: topic, type=topic
"Prepare a research briefing on COP30"	Scaffold: event, type=event

---

Workflow & Vorgehen

1. Name the subject: Extract name/title of the briefing from the user input.
2. Detect type: person, organization, topic, event (or unspecified).
3. Generate scaffold: Create Markdown with all relevant sections.
4. Output: stdout or optionally write to a file (-o file.md).
5. Start research: Hand scaffold to research-agent or web-reading to fill in missing sections.

---

CLI

# Briefing to stdout (Deutsch)
PYTHONDONTWRITEBYTECODE=1 python dossier_briefing_core.py "Marie Curie" --typ person

# Write to file (Deutsch)
PYTHONDONTWRITEBYTECODE=1 python dossier_briefing_core.py "OpenAI" --typ organization -o briefing_openai.md

# Topic briefing (Deutsch)
PYTHONDONTWRITEBYTECODE=1 python dossier_briefing_core.py "Quantum computing" --typ topic

# Event (Deutsch)
PYTHONDONTWRITEBYTECODE=1 python dossier_briefing_core.py "COP30" --typ event

# Without type (generic) (Deutsch)
PYTHONDONTWRITEBYTECODE=1 python dossier_briefing_core.py "My topic"

# Help (Deutsch)
PYTHONDONTWRITEBYTECODE=1 python dossier_briefing_core.py --help

---

Briefing Types and Sections

Type	Sections
person	Basic data, biography/background, work & contributions, sources, notes
organization	Profile, history, products/services, key people, sources, notes
topic	Overview, background/context, current developments, key sources, open questions, notes
event	Key facts, participants, background/timeline, significance, sources, notes
unspecified	Overview, background, details, sources, notes

---

Store

No persistent store. The scaffold is only output (stdout or file), not stored in a database.

---

Attitude

• Always emphasise that the scaffold is empty and must be filled through research.
• Never invent content or hallucinate — only provide structure.
• Ask if the type is unclear or use unspecified.

---

Privacy

No network access. No store. Purely local processing.

---

Related Resources

• research-agent — fills the briefing scaffold with research results
• web-reading — reads web pages and extracts content for the briefing

---

Änderungsprotokoll

Version	Date	Change
1.0.0	2026-06-22	Created from BACH dossier_generator.py v1.0.0; store removed, generalised