Your First Agent

Governance flags

Control what your agent can and can’t do:

from tappass import Agent

agent = Agent("https://tappass.example.com", "tp_...", flags={
    "mode": "observe",                    # log everything, block nothing
    "email": "mirror:you@company.com",    # redirect agent emails to you
    "pii": "mask",                        # mask PII in responses
    "budget": "dev",                      # $1/session cap
})

response = agent.chat("Draft an email to the client about Q4 results")

See Governance Flags for all available flags.

Streaming

for chunk in agent.stream("Write a compliance report"):
    print(chunk, end="", flush=True)

Multi-turn conversations

agent.chat("My name is Alice and I work at Acme Corp")
response = agent.chat("What's my name?")
print(response.content)  # "Your name is Alice."

Async

response = await agent.achat("Analyze the data")

async for chunk in agent.astream("Summarize findings"):
    print(chunk, end="", flush=True)

Error handling

from tappass import PolicyBlockError

try:
    response = agent.chat("Show me all credit card numbers")
except PolicyBlockError as e:
    print(e.blocked_by)       # "detect_pii"
    print(e.reason)           # "PII detected in request"

PolicyBlockError is never retried. It is a deliberate policy decision.