minigist
AI-powered summaries for your Miniflux feeds.
Turn your long Miniflux articles into clear, concise summaries.
✨ Why minigist?
Long-form feed entries are great for depth but hard to keep up with. Miniflux shows the full content, and without a summary layer you still have to read everything to decide what matters.
minigist solves this by generating concise summaries on the server side and writing them directly into each unread entry. This works with your current way to read Miniflux entries, no client changes needed.
🧠 How it works
- You define prompts and targets in a YAML config file
- minigist fetches the full text of unread entries (using pure.md if configured)
- Your configured LLM generates summaries using your prompts
- Summaries are written back into the Miniflux entry
🚀 Installation
Install minigist using pip:
pip install minigist
Install minigist using uv:
uv tool install minigist
📄 Usage
Configuration
Find below an example config file. By default, minigist searches for a config file in this order:
~/.config/minigist/config.yaml~/.config/minigist/config.yml./config.yaml./config.yml/etc/minigist/config.yaml/etc/minigist/config.yml
You can override the path with the --config-file flag.
For credentials, you can set environment variables to override the YAML values:
MINIGIST_MINIFLUX_API_KEY→miniflux.api_keyMINIGIST_LLM_API_KEY→llm.api_key
miniflux:
# Base URL of your Miniflux instance (required)
url: "https://your-miniflux-instance.com"
# Miniflux API key (required)
api_key: "your-miniflux-api-key"
# Request timeout in seconds (optional; default: 2)
timeout_seconds: 2
llm:
# API key for your LLM provider (required)
api_key: "your-ai-service-api-key"
# API base URL (optional; default: OpenRouter)
base_url: "https://openrouter.ai/api/v1"
# Model identifier to use (optional; default shown)
model: "google/gemini-2.5-flash-lite"
# Request timeout in seconds (optional; default: 60)
timeout_seconds: 60
# Max number of concurrent LLM requests (optional; default: 5)
concurrency: 5
prompts:
# Prompts define how summaries are produced.
# Each prompt must have a unique id.
- id: "default"
prompt: "Generate an executive summary of the provided article."
- id: "deep-dive"
prompt: "Extract the nuanced arguments and counterpoints."
# minigist always appends a fixed system instruction that enforces JSON output
# and requires summary_markdown to be Markdown. This cannot be overridden.
# Optional: when no targets are defined, this prompt is used for all unread entries.
# If omitted, the first prompt in the "prompts" list is used.
default_prompt_id: "default"
targets:
# Targets map feeds or categories to prompts.
# When targets are defined, ONLY these feeds/categories are processed.
# Overlaps across targets are errors.
- prompt_id: "default"
feed_ids: [1, 2]
- prompt_id: "deep-dive"
category_ids: [5]
# Prefer pure.md for this target (optional; default: false)
use_pure: true
scraping:
# Token for pure.md (optional; improves rate limits)
pure_api_token: "optional-pure-md-token"
# Always route matching URLs through pure.md
pure_base_urls:
- "https://text.npr.org/"
# Request timeout for scraping in seconds (optional; default: 5)
timeout_seconds: 5
fetch:
# Max unread entries to fetch per feed (optional; default: 50)
limit: 50
notifications:
# Apprise notification URLs for error/failure alerts (optional)
- "discord://webhook_id/webhook_token"
- "telegram://bot_token/chat_id"
See Apprise documentation for all supported notification services.
Run
Run minigist to process unread entries:
minigist run
Run in dry-run mode to see what would happen without making changes:
minigist run --dry-run
Increase logging verbosity:
minigist run --log-level DEBUG
Use a different configuration file:
minigist run --config-file /path/to/config.yaml
Metadata
Release files for minigist 0.9.7
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| minigist-0.9.7.tar.gz | 24.6 kB | Without provenance | Details for minigist-0.9.7.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| minigist-0.9.7-py3-none-any.whl | Python 3 | none | any | Details for minigist-0.9.7-py3-none-any.whl |
Total release size: 54.3 kB
Release files / minigist-0.9.7.tar.gz
| Download URL | minigist-0.9.7.tar.gz |
|---|---|
| Size | 24.6 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
74d0e0e1b2ba63f491ed201a53ee4ef57d2a92dc2b311a7bb9d3664275221627
|
|
BLAKE2b-256 checksum How to use checksums |
525c3c3a0eea5e14b2e2b80d712d7983a60af3f8baee85bfcf23c2ce5cb3c735
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
uv/0.13.6 {"installer":{"name":"uv","version":"0.13.6","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
Release files / minigist-0.9.7-py3-none-any.whl
| Download URL | minigist-0.9.7-py3-none-any.whl |
|---|---|
| Size | 29.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
1a06e558a4bd3ab3f5a2b9d2d4895ff6d51c71e4c0b56b43da79b15c82e5bea6
|
|
BLAKE2b-256 checksum How to use checksums |
f8590edbec6c93aa1020cc986722d5d5733a8cc05b89c13882e2d4b0d7703c79
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
uv/0.13.6 {"installer":{"name":"uv","version":"0.13.6","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
Security for minigist 0.9.7
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.