chore(release): bump version to 3.18.3
This commit is contained in:
@@ -11,7 +11,7 @@
|
||||
{
|
||||
"name": "last30days",
|
||||
"description": "Research any topic across Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, GitHub, and 5+ more sources. AI agent scores by upvotes, likes, and real money - not editors.",
|
||||
"version": "3.18.2",
|
||||
"version": "3.18.3",
|
||||
"author": {
|
||||
"name": "Matt Van Horn",
|
||||
"url": "https://github.com/mvanhorn"
|
||||
|
||||
@@ -1,6 +1,6 @@
|
||||
{
|
||||
"name": "last30days",
|
||||
"version": "3.18.2",
|
||||
"version": "3.18.3",
|
||||
"description": "Research any topic across Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, GitHub, and 5+ more sources. AI agent scores by upvotes, likes, and real money - not editors.",
|
||||
"author": {
|
||||
"name": "Matt Van Horn",
|
||||
|
||||
@@ -1,6 +1,6 @@
|
||||
{
|
||||
"name": "last30days",
|
||||
"version": "3.18.2",
|
||||
"version": "3.18.3",
|
||||
"description": "Research any topic across Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, GitHub, and the web.",
|
||||
"author": {
|
||||
"name": "Matt Van Horn",
|
||||
|
||||
@@ -9,7 +9,7 @@
|
||||
{
|
||||
"name": "last30days",
|
||||
"description": "Research any topic across Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, GitHub, and 5+ more sources. AI agent scores by upvotes, likes, and real money - not editors.",
|
||||
"version": "3.18.2",
|
||||
"version": "3.18.3",
|
||||
"category": "productivity",
|
||||
"source": {
|
||||
"source": "url",
|
||||
|
||||
@@ -1,6 +1,6 @@
|
||||
{
|
||||
"name": "last30days",
|
||||
"version": "3.18.2",
|
||||
"version": "3.18.3",
|
||||
"description": "Research any topic across Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, GitHub, and 5+ more sources. AI agent scores by upvotes, likes, and real money - not editors.",
|
||||
"author": {
|
||||
"name": "Matt Van Horn",
|
||||
|
||||
@@ -9,6 +9,22 @@ This project uses [towncrier](https://towncrier.readthedocs.io/). Upcoming notes
|
||||
|
||||
<!-- towncrier release notes start -->
|
||||
|
||||
## [3.18.3] - 2026-07-25
|
||||
|
||||
### Fixed
|
||||
|
||||
- Top Community Comments ranking blends thread relevance with vote strength and applies the relevance floor only when enough on-topic candidates exist. ([#701](https://github.com/mvanhorn/last30days-skill/issues/701))
|
||||
- Chromium cookie extraction now searches every browser profile for a matching cookie set, and reuses the Keychain/AES key across profiles in one scan. ([#725](https://github.com/mvanhorn/last30days-skill/issues/725))
|
||||
- Synthesis contract is echoed at the top of the evidence envelope so hosts that truncate stdout still see the "synthesize, don't dump" directive. ([#727](https://github.com/mvanhorn/last30days-skill/issues/727))
|
||||
- `store_findings` no longer raises `TypeError` when a re-sighted finding carries `engagement_score: None`. ([#796](https://github.com/mvanhorn/last30days-skill/issues/796))
|
||||
- Explicit --plan payloads with an invalid schema now exit with a field-specific error instead of silently running a deterministic plan. ([#841](https://github.com/mvanhorn/last30days-skill/issues/841))
|
||||
- Company-topic runs no longer auto-add the `jobs` source when an explicit `--search` / `requested_sources` filter is set; `--hiring-signals` still forces jobs. ([#842](https://github.com/mvanhorn/last30days-skill/issues/842))
|
||||
- Report footer path now matches the collision-safe path actually reserved by save_output. ([#850](https://github.com/mvanhorn/last30days-skill/issues/850))
|
||||
- YouTube ScrapeCreators transcript rescue is logged instead of being masked as a hard failure. ([#851](https://github.com/mvanhorn/last30days-skill/issues/851))
|
||||
- Polymarket domain-sweep topics no longer drop every market after noise-word stripping removes terms like "ai". ([#859](https://github.com/mvanhorn/last30days-skill/issues/859))
|
||||
- Reddit fetch windows now track the requested date range so short `--days` runs no longer pull a depth-default month and discard everything outside the window. ([#860](https://github.com/mvanhorn/last30days-skill/issues/860))
|
||||
|
||||
|
||||
## [3.18.2] - 2026-07-25
|
||||
|
||||
### Added
|
||||
|
||||
@@ -1 +0,0 @@
|
||||
Top Community Comments ranking blends thread relevance with vote strength and applies the relevance floor only when enough on-topic candidates exist.
|
||||
@@ -1 +0,0 @@
|
||||
Chromium cookie extraction now searches every browser profile for a matching cookie set, and reuses the Keychain/AES key across profiles in one scan.
|
||||
@@ -1 +0,0 @@
|
||||
Synthesis contract is echoed at the top of the evidence envelope so hosts that truncate stdout still see the "synthesize, don't dump" directive.
|
||||
@@ -1 +0,0 @@
|
||||
`store_findings` no longer raises `TypeError` when a re-sighted finding carries `engagement_score: None`.
|
||||
@@ -1 +0,0 @@
|
||||
Explicit --plan payloads with an invalid schema now exit with a field-specific error instead of silently running a deterministic plan.
|
||||
@@ -1 +0,0 @@
|
||||
Company-topic runs no longer auto-add the `jobs` source when an explicit `--search` / `requested_sources` filter is set; `--hiring-signals` still forces jobs.
|
||||
@@ -1 +0,0 @@
|
||||
Report footer path now matches the collision-safe path actually reserved by save_output.
|
||||
@@ -1 +0,0 @@
|
||||
YouTube ScrapeCreators transcript rescue is logged instead of being masked as a hard failure.
|
||||
@@ -1 +0,0 @@
|
||||
Polymarket domain-sweep topics no longer drop every market after noise-word stripping removes terms like "ai".
|
||||
@@ -1 +0,0 @@
|
||||
Reddit fetch windows now track the requested date range so short `--days` runs no longer pull a depth-default month and discard everything outside the window.
|
||||
@@ -1,6 +1,6 @@
|
||||
{
|
||||
"name": "last30days-skill",
|
||||
"version": "3.18.2",
|
||||
"version": "3.18.3",
|
||||
"description": "Research a topic from the last 30 days across Reddit, X, YouTube, TikTok, Instagram, Hacker News, Polymarket, and the web.",
|
||||
"settings": [
|
||||
{
|
||||
|
||||
+1
-1
@@ -1,6 +1,6 @@
|
||||
[project]
|
||||
name = "last30days-skill"
|
||||
version = "3.18.2"
|
||||
version = "3.18.3"
|
||||
description = "Multi-source last-30-days research skill"
|
||||
readme = "README.md"
|
||||
requires-python = ">=3.12"
|
||||
|
||||
@@ -1,6 +1,6 @@
|
||||
---
|
||||
name: last30days
|
||||
version: "3.18.2"
|
||||
version: "3.18.3"
|
||||
description: "Research what people actually say about any topic in the last 30 days. Pulls posts and engagement from Reddit, X, YouTube, TikTok, Hacker News, Polymarket, GitHub, and the web. Includes a doctor health check to diagnose broken or missing sources."
|
||||
argument-hint: 'last30days nvidia earnings reaction | last30days AI video tools | last30days what users want in react'
|
||||
allowed-tools: Bash, Read, Write, AskUserQuestion, WebSearch
|
||||
@@ -412,7 +412,7 @@ If your Bash call to `last30days.py` does NOT include the FULL pre-flight checkl
|
||||
|
||||
---
|
||||
|
||||
# last30days v3.18.2: Research Any Topic from the Last 30 Days
|
||||
# last30days v3.18.3: Research Any Topic from the Last 30 Days
|
||||
|
||||
> **Permissions overview:** Reads public web/platform data and optionally saves research briefings to `LAST30DAYS_MEMORY_DIR` (defaults to `~/Documents/Last30Days`). X/Twitter search uses optional user-provided tokens (AUTH_TOKEN/CT0 env vars). Bluesky search uses optional app password (BSKY_HANDLE/BSKY_APP_PASSWORD env vars - create at bsky.app/settings/app-passwords). On hosts with `uv` and no Python 3.12+, the preflight may install a uv-managed CPython 3.12 (one-time ~28MB download, announced on stderr). All credential usage and data writes are documented in the [Security & Permissions](#security--permissions) section.
|
||||
|
||||
|
||||
Reference in New Issue
Block a user