Configure feed alert and digest thresholds via distillery_configure with validation.
npx skills add https://github.com/norrietaylor/distillery --skill tune-norrietaylor
Please help me install this Agent Skill. Skill: tune Source: https://github.com/norrietaylor/distillery/tree/main/skills/tune Command: npx skills add https://github.com/norrietaylor/distillery --skill tune-norrietaylor
Managing feed relevance thresholds manually can lead to missed priorities or too much noise. This skill provides a centralized way to view and adjust the alert and digest thresholds that control what items trigger alerts and what items are included in digests.
Tune the feed thresholds by specifying the alert and digest values using the available flags.
High-intent search queries and answers about installing and using this skill.
To reduce alert noise, raise the alert threshold using the available flags. A higher alert threshold ensures only highly relevant items trigger alerts, filtering out lower-priority feed items.
Valid alert and digest threshold values range from 0.0 to 1.0. Input validation enforces this range and also requires the alert threshold to be greater than or equal to the digest threshold.
You are missing digest items because the threshold is too high. Lower the digest threshold to include more feed items with lower relevance scores in your generated digests.
Yes, you can reset alert and digest thresholds to their default values. The skill supports a reset function to restore default configurations across interactive sessions and automated workflows.
Yes, the threshold tuning process includes a confirmation step before applying changes. It enforces runtime consistency by applying validated configurations via distillery_configure after confirmation.
TravisLeeeeeeScan news sources, filter by relevance, and generate concise AI-focused summaries.
Leep-GHRoute agent notifications to channels by type using CHANNEL tags.
nssuwan186-devRoute execution to character-specific bash scripts via JSON configuration and CLI arguments.
OmniNode-aiGenerate prioritized work suggestions from a Linear backlog by filtering blocked issues.
