Help us improve
Share bugs, ideas, or general feedback.
From runesmith-devtools
Propagate convention changes across skills safely by orchestrating per-skill modifications through Anthropic's skill-creator. Use after a decision changes how the system works - variable renames, workflow changes, lib reference updates, removed features, new conventions. Triggers on: "update skills", "propagate this", "skills are stale", "sync the change", "make sure skills reflect this", or when a decision contradicts what existing skills describe.
npx claudepluginhub runemarklabs/runesmith --plugin runesmith-devtoolsHow this skill is triggered — by the user, by Claude, or both
Slash command
/runesmith-devtools:skill-updaterThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Orchestrator for cross-skill convention propagation. This skill is the **coordinator**: it knows which skills to touch and what change to apply. Anthropic's `skill-creator` skill is the **executor**: it knows how to safely modify one skill at a time, with frontmatter validation and trigger-phrase eval.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Guides Next.js Cache Components and Partial Prerendering (PPR): 'use cache' directives, cacheLife(), cacheTag(), revalidateTag() for caching, invalidation, static/dynamic optimization. Auto-activates on cacheComponents: true.
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Share bugs, ideas, or general feedback.
Orchestrator for cross-skill convention propagation. This skill is the coordinator: it knows which skills to touch and what change to apply. Anthropic's skill-creator skill is the executor: it knows how to safely modify one skill at a time, with frontmatter validation and trigger-phrase eval.
This split exists so that bulk changes never corrupt skill files. Every per-skill rewrite goes through skill-creator's modify path and gets validated before the next skill is touched.
lib/install-paths.md - runtime detection and plugin scan pathsskill-creator skill (ships with Claude) - used for per-skill modify + validationlib/user-prompts.md - structured-input requirement for any user promptThis skill follows the marketplace-wide user-prompt standard in lib/user-prompts.md. Every user prompt MUST use the host client's structured input UI (single-pick, multi-pick, or text-input form). Never freeform plain-text yes/no questions. The only exception is the consent-trigger gate documented in lib/consent.md, which waits for user-initiated phrases like "make the ticket".
Per install-paths.md:
CLAUDE_PLUGINS_DIR env var if set~/.claude/plugins/ (Claude Code / Teams)rpm/plugin_*/plugins/*/ in this repoIf none found: "No plugins installed and no marketplace source. Nothing to update."
skill-creator is an Anthropic-shipped skill available in any Claude environment. Verify it can be invoked. If somehow unavailable: degrade to dry-run mode (preview-only, no writes).
Use for:
{SPACE_KEY} → {CONFLUENCE_SPACE})Do not use for:
skill-creator directlyAsk user:
Reject vague changes ("update the skills"). Demand specificity (e.g. "rename any reference from one library file name to another across all runesmith plugins, with the exact old and new names provided").
Scan plugin source per install-paths.md. For each SKILL.md:
Build a list:
Affected: 12 SKILL.md files across 4 plugins
runesmith-jira/skills/ticket/SKILL.md (3 occurrences)
runesmith-jira/skills/bug-report/SKILL.md (1 occurrence)
runesmith-confluence/skills/feature-doc/SKILL.md (2 occurrences)
...
For each affected file, show the change preview:
name:, description:, or model:)Ask: "Apply these