agentleFS
anmoba99anmoba99/pipedrive-claude-skills/skills/pd-setup/SKILL.md

pd-setup

Skill0 starsChanged 29 days agoMITView on GitHub

One-time setup and health check ("doctor") for the pd-* Pipedrive skills. Caches pipelines, stages, users and custom fields, records industry and output preferences, and reports which skills are ready. Use on first install, when something breaks, or after adding a connector.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Follow pd-core rules (load pd-core if not loaded).

    line 11
  2. Very specific:

    Account context: Call getAccountContext once.

    line 20
  3. Very specific:

    Industry check: If an industry is chosen, load pd-industry/references/<industry>.md.

    line 42
  4. Very specific:

    Access check: Run pd-access checks 1 to 4 (connector permissions, coverage, read-only fields, role fit) and save .pipedrive/access.json.

    line 43

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

pd-setup

Follow pd-core rules (load pd-core if not loaded).

Requires: Pipedrive Optional: none Fallback: no filesystem (claude.ai chat) -> keep context in this conversation only and say so.

Steps

  1. Connectors. List which integrations are live in this session by checking available tools: Pipedrive, Slack, Teams, Gmail, Outlook, Google Drive, Google Calendar. Pipedrive missing: stop and say "Connect Pipedrive in Settings > Connectors (or claude mcp add in Claude Code), then run pd-setup again."
  2. Account context. Call getAccountContext once.
  3. Option labels. Build field-map.json from the account context first: each custom field has field_name, field_code, field_type, options (id, label) and is_writable. Use the export below only as a fallback: when option fields come back with empty options, or the user wants a copy in a Claude Project. Look for the user's export: Project files, a chat attachment, or .pipedrive/data_fields_options_*.csv. It is a CSV with columns Option ID, Option name, Item type, Field name, Field type, Field API key, one row per option, single and multiple option fields only (other field types are not in it; take those from getAccountContext).
    • Found, shell available: python3 scripts/field_options.py <export.csv> --merge .pipedrive/field-map.json --out .pipedrive/field-map.json after step 4 writes the base map. Pass the duplicate-name warning from stderr to the report.
    • Found, no shell: read it and fill options in field-map.json (or keep in conversation). Match by Field API key, never by name: names can repeat and have trailing spaces.
    • Not found and option labels are missing from the context: ask once, "In Pipedrive go to Settings > Data fields, click Export, choose Selectable options in fields, and upload the CSV (add it to a Claude Project so every chat has it)." Continue without it if the user skips.
  4. Write cache (if filesystem available) to .pipedrive/ in the working folder:
    • context.json: slim version only:
      {"fetched": "<ISO date>", "me": {"id": 0, "name": ""},
       "pipelines": {"<id>": {"name": "", "stages": {"<id>": ""}}},
       "stages": {"<id>": "<name>"}, "users": {"<id>": "<name>"}}
    • account_context.json: the raw getAccountContext result, unchanged. Scripts read stage order, custom fields, option labels and is_writable from it; the slim file above lacks them.
    • field-map.json: {"deal": {"<Field name>": {"key": "<hash>", "type": "", "options": {"<label>": <id>}}}, "person": {}, "org": {}}
    • config.json: from references/config.template.json, filled with answers from step 5.
    • last_run.json: {}
  5. Ask once (one message, skip what is known):
    • Industry (offer the pd-industry list, or "generic").
    • Your role: rep, manager, owner, ops, CS.
    • Where team posts go, if Slack/Teams is live (channel name or "chat only").
    • Stale threshold in days (default 14).
    • "Do you manage existing customers in Pipedrive (account management, renewals, upsell)?" On yes, map accounts in config.json from the cached fields: account manager, tier, revenue, renewal date, health score (optional write-back), person role. Follow pd-account-health/references/setup.md, show the mapping, confirm. Missing fields go to the manual steps in step 8.
  6. Industry check. If an industry is chosen, load pd-industry/references/<industry>.md. Compare its "Custom fields worth adding" and pipeline template with the real account. List what is missing as manual steps (the connector cannot create fields or stages).
  7. Access check. Run pd-access checks 1 to 4 (connector permissions, coverage, read-only fields, role fit) and save .pipedrive/access.json. If the context errors show missing permissions, me or users stay empty in context.json; pd-access fills me from the user's own deals.
  8. Report in this format:
Setup done. Cache: .pipedrive/ (fields: 34, pipelines: 2, users: 6)
Option fields: labels from Pipedrive (34) | missing, upload "Selectable options in fields" export
Access: full | partial, 4 of 6 owners visible | connector missing users:read (run pd-access for fixes)
Connectors: Pipedrive live | Slack live | Gmail missing | Calendar missing
Ready: pd-morning-brief, pd-follow-up, pd-pipeline-review, ...
Fallback: pd-follow-up -> drafts in chat (connect Gmail for drafts)
Manual Pipedrive steps for <industry>:
1. Add deal field "Trial end date" (date)
2. ...
Try: "brief me for today"

Doctor mode

User says "doctor", "something broke" or a skill hit a field/stage error: re-run steps 1 to 4 (overwrite cache), keep config.json, report only what changed.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.