3
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill cleanexpo/ato --skill idea-queue-capture- SKILL.md11.5 KB
Overview
This skill instantly captures user ideas into a shared work queue with intelligent grouping so teams can preserve context without interrupting current work. It focuses on speed: capture the user message verbatim, classify type, generate a concise title, and enqueue items for later validation. The capture step avoids planning, deduplication, or execution.
How this skill works
The skill parses user input to extract distinct ideas, detects lists and sentence boundaries, and preserves the full original message as the queue item description. It applies grouping rules to merge related ideas into one item or split unrelated ideas into multiple items, auto-detects a queue_item_type from keywords, generates a concise title (max 80 chars), and writes the item to the work_queue with metadata and timestamp. A short confirmation message returns counts, titles, types, and queue positions.
When to use it
- User provides a feature request, bug report, or improvement idea
- User mentions multiple related ideas that should be grouped
- You want to queue work without interrupting current tasks
- Quick capture is needed without planning or validation overhead
- User issues explicit capture commands like /capture-idea or /capture-request
Best practices
- Preserve user wording verbatim in the description to retain context for downstream validation
- Use clear separators (bullets, numbers, “Also…”, “Separately…”) when you intend separate queue items
- Keep capture fast—avoid adding acceptance criteria or execution steps at capture stage
- Warn users if messages include sensitive data and advise review before validation
- Prefer grouping when ideas affect the same component or have dependencies
Example use cases
- User reports two UI issues affecting the same panel; capture as one grouped queue item
- User verbally lists a bug, a feature, and a compliance analysis; capture as three separate items
- Use /capture-request to enqueue a specific suggestion from a chat thread
- Run /capture-multiple when you explicitly want each idea separated into its own queue item
- Call /list-queue to review pending positions after capture
FAQ
No. Deduplication is deferred to PM validation. Capture is intentionally fast and preserves full context for later review.
What happens if the database write fails?
The skill logs the error, returns a user-friendly message with troubleshooting suggestions, and attempts a local fallback so the user idea is not lost.