Cerb Docs

Cerb Docs

1011 Cerb documentation pages stored as local Markdown files under references/. Each file has YAML frontmatter (id, title, url, summary, tags) followed by the full page content.

How to Search

Grep the references directory for keywords, then read the matching files:

# Find files mentioning a topic
grep -rl "KEYWORD" cerb-docs/references/

# Search within a subtree
grep -rl "oauth" cerb-docs/references/guides/integrations/

# Search titles/summaries only (frontmatter)
grep -l "title:.*KEYWORD" cerb-docs/references/**/*.md

# Case-insensitive content search with context
grep -ri "webhooks" cerb-docs/references/docs/ -l

Read matched files in full — they contain complete documentation content.

Doc Gaps — Proposing New or Improved Documentation

When a question can't be adequately answered from the local references, check GitHub for an existing proposal before filing a new one.

1. Check for existing issues (no auth required):

gh issue list --repo cerb/cerb.github.io --search "KEYWORDS" --state open

Search with 2-3 specific keywords from the question. If a matching open issue exists, note it and do not create a duplicate.

2. If no existing issue, create one (requires gh auth login):

gh issue create --repo cerb/cerb.github.io \
  --label "ai-proposed" \
  --title "Docs gap: CONCISE TITLE" \
  --body "$(cat <<'EOF'
## Question
[The user's question verbatim or paraphrased]

## Gap
[What's missing or inadequate in the current docs — be specific]

## Closest existing page
[URL of the nearest existing doc, or "none found"]

## Suggested content
[Brief outline of what the new/improved page should cover]
EOF
)"

Always include --label "ai-proposed". The title must start with Docs gap: so automated tooling can identify it. Omit the issue body section ## Suggested content if you have nothing useful to add.

If gh is not authenticated, surface the gap in conversation and suggest the user visit https://github.com/cerb/cerb.github.io/issues/new to file it manually.

Never Edit Reference Files Directly

The references/ directory is auto-generated from the official Cerb documentation at https://cerb.ai/docs/. Do not edit any file under references/ by hand — changes will be overwritten the next time the fetch tool runs.

To improve or correct documentation: use the Doc Gaps workflow above to file an issue on cerb/cerb.github.io. That is the authoritative source.

Updating the Reference Files

Run the fetch tool to pull the latest docs from cerb.ai:

python3 cerb-docs/tools/fetch-docs.py --clean
# or from a local JSONL snapshot:
python3 cerb-docs/tools/fetch-docs.py --input /path/to/search.jsonl --clean

Related Skills

These skills are always installed alongside this one:

• /cerb-automations — write or modify Cerb automations (KATA, commands, triggers, events)
• /cerb-dev — make changes to the Cerb PHP/MySQL codebase (new record types, extensions, migrations)
• /cerb-search — build search queries for any record type

Directory Structure

references/
  docs/
    api/
      authentication/
      endpoints/
      libraries/
      records/
    automations/
      commands/
      events/
      inputs/
      triggers/
        interaction.website/
          elements/
        interaction.worker/
          elements/
    dashboards/
      widgets/
    data-queries/
      attachment/
      autocomplete/
      automation/
      calendar/
      classifier/
      data/
      gpg/
      metrics/
      platform/
      record/
      ui/
      usage/
      worklist/
    developers/
    groups/
    guide/
      admins/
      developers/
    installation/
    plugins/
    portals/
    records/
      fields/
        types/
      types/
    scripting/
      filters/
      functions/
    setup/
      configure/
      developers/
      mail/
      packages/
      plugins/
      records/
      storage/
      team/
    toolbars/
      interactions/
  guides/
    ai-agents/
    api/
    automations/
      automation.timer/
      interaction.worker/
    bots/
    classifiers/
    developers/
    impex/
    installation/
    integrations/
      aws/
      azure/
      google/
      jotform/
      salesforce/
      slack/
    interactions/
      website/
    localization/
    machine-learning/
      image-generation/
      question-answering/
    mail/
      encryption/
      privacy/
    packages/
    portals/
    project-boards/
    records/
    scaling/
    security/
    sso/
    webhooks/
  releases/
  resources/
  solutions/
    automations/
    integrations/
  tips/
  workflows/

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: cerb
• Source: cerb/claude-skills
• License: MIT

Install and usage instructions live in the source repository linked above.