Claude AI for Collaborative Work — Complete Guide: Teams, Projects, and Workflows

System: You are a research analyst. Be precise, cite uncertainty when present,
and structure your analysis clearly. When you don't know something, say so.

User:
I'm attaching [document/report/data]. Please:

1. Executive Summary (3-4 bullets, max 100 words)
2. Key findings with supporting evidence from the document
3. Gaps or limitations in the data or analysis
4. Implications for [your specific context/decision]
5. Questions I should investigate further

Focus specifically on [aspect]. Note any assumptions you're making.
Avoid hedging everything — be direct about what the data shows.

[Paste document content]

Review this code for the following, in order of priority:

1. Security vulnerabilities (SQL injection, XSS, insecure direct object reference, auth bypass)
2. Performance issues (N+1 queries, missing indexes, blocking operations)
3. Error handling gaps (unhandled exceptions, missing validation, silent failures)
4. Code quality and maintainability (clarity, duplication, overly complex logic)
5. Test coverage suggestions

For each issue found, provide:
- Severity: Critical / High / Medium / Low
- Location: file:line
- Issue: clear description of the problem
- Risk: what could go wrong if not fixed
- Suggested fix: working code example

Flag anything that looks like a potential race condition or concurrency issue.

[Paste code]

import anthropic
from pathlib import Path

client = anthropic.Anthropic()

def review_pull_request(changed_files: list[dict]) -> str:
    """
    Review a set of changed files from a pull request.
    changed_files: [{"filename": "...", "patch": "...", "content": "..."}]
    """
    files_text = "

".join(
        f"### {f['filename']}
```
{f['content']}
```"
        for f in changed_files
    )

    message = client.messages.create(
        model="claude-3-5-sonnet-20241022",
        max_tokens=4096,
        system=(
            "You are a senior engineer conducting a thorough code review. "
            "Be specific and actionable. Provide code examples for all suggestions."
        ),
        messages=[{
            "role": "user",
            "content": f"Review these changed files:\n\n{files_text}"
        }]
    )
    return message.content[0].text

# Use in a GitHub Actions workflow
# result = review_pull_request(get_pr_files(pr_number))
# post_pr_comment(pr_number, result)

import anthropic

client = anthropic.Anthropic()

def analyze_document(document_text: str, analysis_type: str) -> str:
    """Analyze a document using Claude's 200K context window."""
    message = client.messages.create(
        model="claude-3-5-sonnet-20241022",
        max_tokens=4096,
        system="You are a business analyst. Provide structured, actionable analysis.",
        messages=[
            {
                "role": "user",
                "content": f"Analyze this {analysis_type}:\n\n{document_text}"
            }
        ]
    )
    return message.content[0].text


def batch_analyze(documents: list[tuple[str, str]]) -> list[dict]:
    """
    Analyze multiple documents.
    documents: [(content, type), ...]
    """
    results = []
    for content, doc_type in documents:
        analysis = analyze_document(content, doc_type)
        results.append({"type": doc_type, "analysis": analysis})
    return results


# Use in a Slack bot, web app, or internal tool
# result = analyze_document(contract_text, "vendor contract for SaaS tool")
# post_to_slack(result)

# Streaming for long documents:
with client.messages.stream(
    model="claude-3-5-sonnet-20241022",
    max_tokens=4096,
    messages=[{"role": "user", "content": "Analyze this 100-page report: ..."}],
) as stream:
    for text_chunk in stream.text_stream:
        print(text_chunk, end="", flush=True)  # stream to frontend in real-time