From refinery
Use when searching documentation about a concept across platforms. Use when user says 'search docs', 'buscar documentación', 'analyze docs', 'find documentation', or needs to understand existing documentation about a topic.
How this skill is triggered — by the user, by Claude, or both
Slash command
/refinery:analyze-docssonnetThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Read `${CLAUDE_PLUGIN_ROOT}/config.yaml`.
Read ${CLAUDE_PLUGIN_ROOT}/config.yaml.
If it does not exist, tell the user to copy config.example.yaml to config.yaml and fill in their values. Stop here.
Extract:
docs_sources — array of documentation sources (each with type and optional cloud_id)If docs_sources is empty or not defined, tell the user no documentation sources are configured and stop.
For each source in {config.docs_sources}:
${CLAUDE_PLUGIN_ROOT}/skills/analyze-docs/adapters/{source.type}.mdStore all loaded adapters and their associated source configs.
From $ARGUMENTS:
For each configured documentation source, use the adapter's search command to find relevant pages.
Search strategy:
Collect the top 5 results per source.
For each relevant page found:
Critical rule: Extract only what the documentation explicitly states. Do not infer, summarize, or add context that is not in the source.
From the extracted content, identify explicit business rules:
For each rule, record:
Compare information across sources. Flag:
For each aspect of the search topic, assess documentation coverage:
| Aspecto | Fuente 1 | Fuente 2 | ... | Cobertura |
|---|---|---|---|---|
| {aspect} | ✅ Documented | ❌ Not found | ... | {partial/full/none} |
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
📚 DOCUMENTATION ANALYSIS: "{search_term}"
Sources: {count} | Pages found: {total}
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
━━ SOURCE: {source_name} ({source_type}) ━━━━━━
📄 {page_title}
URL: {url}
Updated: {date}
Relevant content:
> "{direct quote 1}"
> "{direct quote 2}"
📄 {page_title_2}
...
━━ BUSINESS RULES FOUND ━━━━━━━━━━━━━━━━━━━━━━
1. "{exact rule}" — Source: {page_title} ({source})
2. "{exact rule}" — Source: {page_title} ({source})
━━ CONTRADICTIONS ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚠️ {source_a}/{page_a} says: "{X}"
{source_b}/{page_b} says: "{Y}"
→ Needs clarification from: {audience}
━━ COVERAGE ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
{coverage table from Step 7}
━━ NOT DOCUMENTED ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
The following aspects were NOT found in any source:
- {missing_aspect_1}
- {missing_aspect_2}
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
npx claudepluginhub weorbitant/workbench-dev --plugin refineryFetches up-to-date documentation from Context7 for libraries and frameworks like React, Next.js, Prisma. Use for setup questions, API references, and code examples.
Applies a firm's KYC/AML rules grid to parsed onboarding records: assigns risk rating, checks required documents, outputs rule outcomes with citations, and routes for escalation.
Generates daily or weekly digests of activity from connected sources (chat, email, docs, tasks, CRM), highlighting action items, decisions, mentions, and project updates.