OpenParable
Skip to content
← Back to skills

Palantir SDK Patterns

Development
palantir-sdk-patterns

Palantir SDK Patterns is an Agent Skill for development published in the jeremylongshore/claude-code-plugins-plus-skills repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Apply production-ready Palantir Foundry SDK patterns for Python and TypeScript. Use when implementing Foundry integrations, refactoring SDK usage, or establishing team coding standards for Foundry API calls. Trigger with phrases like "palantir SDK patterns", "foundry best practices", "palantir code patterns", "idiomatic foundry SDK".

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/jeremylongshore/claude-code-plugins-plus-skills/blob/HEAD/plugins/saas-packs/palantir-pack/skills/palantir-sdk-patterns/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/palantir-sdk-patterns, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Palantir SDK Patterns

Overview

Production-ready patterns for Foundry Platform SDK and OSDK usage. Covers client singletons, typed error handling, pagination helpers, retry logic, and multi-tenant client factories.

Prerequisites

  • Completed palantir-install-auth setup
  • Familiarity with async/await patterns
  • foundry-platform-sdk or @osdk/client installed

Instructions

Step 1: Singleton Client (Python)

# src/foundry_client.py
import os
import foundry
from functools import lru_cache

@lru_cache(maxsize=1)
def get_client() -> foundry.FoundryClient:
    """Thread-safe singleton — cached after first call."""
    auth = foundry.ConfidentialClientAuth(
        client_id=os.environ["FOUNDRY_CLIENT_ID"],
        client_secret=os.environ["FOUNDRY_CLIENT_SECRET"],
        hostname=os.environ["FOUNDRY_HOSTNAME"],
        scopes=["api:read-data", "api:write-data"],
    )
    auth.sign_in_as_service_user()
    return foundry.FoundryClient(auth=auth, hostname=os.environ["FOUNDRY_HOSTNAME"])

Step 2: Typed Error Handling

import foundry
from dataclasses import dataclass
from typing import TypeVar, Generic, Optional

T = TypeVar("T")

@dataclass
class Result(Generic[T]):
    data: Optional[T] = None
    error: Optional[str] = None
    status_code: Optional[int] = None

def safe_call(fn, *args, **kwargs) -> Result:
    """Wrap any Foundry SDK call with structured error handling."""
    try:
        return Result(data=fn(*args, **kwargs))
    except foundry.ApiError as e:
        return Result(error=e.message, status_code=e.status_code)
    except Exception as e:
        return Result(error=str(e))

# Usage
result = safe_call(
    get_client().ontologies.OntologyObject.list,
    ontology="my-company", object_type="Employee", page_size=10,
)
if result.error:
    print(f"Error {result.status_code}: {result.error}")
else:
    print(f"Found {len(result.data.data)} objects")

Step 3: Pagination Helper

def paginate_objects(client, ontology: str, object_type: str, page_size: int = 100):
    """Iterate through all objects with automatic pagination."""
    page_token = None
    while True:
        result = client.ontologies.OntologyObject.list(
            ontology=ontology,
            object_type=object_type,
            page_size=page_size,
            page_token=page_token,
        )
        yield from result.data
        page_token = result.next_page_token
        if not page_token:
            break

# Usage
for emp in paginate_objects(get_client(), "my-company", "Employee"):
    print(emp.properties["fullName"])

Step 4: Retry with Exponential Backoff

import time
import random

def retry_with_backoff(fn, max_retries=3, base_delay=1.0):
    """Retry on 429/5xx with jittered exponential backoff."""
    for attempt in range(max_retries + 1):
        try:
            return fn()
        except foundry.ApiError as e:
            if attempt == max_retries or e.status_code not in (429, 500, 502, 503):
                raise
            delay = base_delay * (2 ** attempt) + random.uniform(0, 0.5)
            print(f"Retry {attempt + 1}/{max_retries} in {delay:.1f}s (HTTP {e.status_code})")
            time.sleep(delay)

Step 5: TypeScript OSDK Patterns

import { createClient, type Client } from "@osdk/client";
import { createConfidentialOauthClient } from "@osdk/oauth";

// Singleton with lazy initialization
let _client: Client | null = null;

export function getOsdkClient(): Client {
  if (!_client) {
    const oauth = createConfidentialOauthClient(
      process.env.FOUNDRY_CLIENT_ID!,
      process.env.FOUNDRY_CLIENT_SECRET!,
      `https://${process.env.FOUNDRY_HOSTNAME}/multipass/api/oauth2/token`,
    );
    _client = createClient(
      `https://${process.env.FOUNDRY_HOSTNAME}`,
      process.env.ONTOLOGY_RID!,
      oauth,
    );
  }
  return _client;
}

Output

  • Thread-safe singleton client with cached auth
  • Structured Result type for error handling
  • Automatic pagination for large object sets
  • Retry logic with jittered backoff

Error Handling

PatternUse CaseBenefit
SingletonAll API callsOne auth flow, reused connection
Result wrapperError propagationNo uncaught exceptions
Pagination helperLarge datasetsMemory-safe iteration
Retry with backoffTransient failuresResilient operations

Examples

Multi-Tenant Client Factory

_clients: dict[str, foundry.FoundryClient] = {}

def get_client_for_tenant(tenant_id: str) -> foundry.FoundryClient:
    if tenant_id not in _clients:
        hostname = get_tenant_hostname(tenant_id)
        token = get_tenant_token(tenant_id)
        _clients[tenant_id] = foundry.FoundryClient(
            auth=foundry.UserTokenAuth(hostname=hostname, token=token),
            hostname=hostname,
        )
    return _clients[tenant_id]

Resources

Next Steps

Apply patterns in palantir-core-workflow-a for real pipeline usage.