metaskill-triggering_skill

This skill optimizes trigger keywords and frontmatter description to reliably activate the metaskill, ensuring specificity and keyword-variants.
  • Shell

1

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 gigaverse-app/skillet --skill metaskill-triggering

  • SKILL.md5.4 KB

Overview

This skill optimizes skill triggers and descriptions to ensure reliable automatic activation. I focus on crafting a single, keyword-rich description that Claude reads at startup and uses for discovery. The goal is predictable triggering by combining what the skill does, specific actions, when to use it, and the trigger keywords.

How this skill works

At startup the agent loads only the name and description for discovery, so the description alone determines auto-triggering. I inspect the description for four parts: what it does, specific actions, when to use it, and a comprehensive set of trigger keywords and phrases. I also debug common failures like vague descriptions, missing variants, lacking "use when" clauses, or disabled model invocation.

When to use it

  • Skill never auto-triggers or triggers inconsistently
  • You want to add or refine trigger keywords and user phrases
  • Writing or editing a skill description or frontmatter
  • Debugging skill activation during agent planning
  • Preparing skills for predictable, low-noise discovery

Best practices

  • Make the description explicit: state what the skill does and specific actions it performs
  • Include an explicit "Use when" clause describing conditions or planned agent actions
  • List keyword variants, synonyms, slang, and user phrases (e.g., "test", "tests", "testing", "add tests")
  • Prefer action verbs and short phrases over passive nouns
  • Keep description concise and keyword-dense (aim under ~1024 characters)
  • Verify disable-model-invocation is not set to true, as that blocks discovery

Example use cases

  • Convert a vague description into the trigger formula so it reliably fires on user prompts
  • Add casual user phrases and abbreviations to catch conversational requests
  • Debug a skill that never runs by checking description presence and keyword coverage
  • Add a "when to use" clause so the agent triggers the skill while planning file edits
  • Test trigger behavior with a trigger-tester agent to validate coverage

FAQ

Only the description field affects auto-triggering; other fields affect execution but not discovery.

How do I avoid false positives?

Make the description specific about scope and language (e.g., target language or context) and include precise "use when" conditions to reduce noise.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational