vv-agent

中文文档

A lightweight agent framework extracted from VectorVein's production runtime. Cycle-based execution with pluggable LLM backends, tool dispatch, memory compression, and durable session scheduling.

Install

contract.lock.json pins this repository's language-neutral contract. The schema-2 central support matrix records adoption and verification: Python is the required implementation; vv-agent-rs is frozen at contract 23.0.0 / package series 0.21.x for maintenance only. Python adoption does not require a Rust update. See the contract workflow. This repository keeps a Python-idiomatic API.

python -m pip install -e .

Install the postgres extra for PostgreSQL SessionStore or the s3 extra for S3 workspace storage. Repository HEAD is forward-only: current readers accept only the current strict public and wire shapes.

For both extras: python -m pip install 'vv-agent[postgres,s3]'. See the 0.22.0 release notes for breaking changes.

Current HEAD uses contract v24, public API v8 and one session kernel execution path. Older runtime behavior is retained in Git tags. See v8 migration for API replacements and host seed examples. See cloud host integration for PostgreSQL transactions, queue dispatch and a host-owned Celery example.

Architecture

Runner / InteractiveAgentClient / CLI / AppServer
  -> SessionDriver -> SessionStore (SQLite or PostgreSQL)
  -> session.kernel.drive
  -> retained model/tool operations and child delivery
  -> RunResult / RunEvent / tracing / protocol projections

The public SDK entry points are exported from vv_agent: Agent, Runner, RunConfig, RunHandle, ModelSettings, function_tool, SessionStore, PromptBundle, PromptSection, ToolExecutionResult, ToolArtifactRef, ToolResultCursor, typed RunEvent objects, ApprovalProvider, ContextProvider, SessionRunEventStore, and the interactive session API for desktop/runtime integrations. Extension points that live in package modules include vv_agent.memory.MemoryProvider and vv_agent.tools.ToolExecutor. Lower-level runtime implementation details include AgentTask, AgentResult, Message, CycleRecord, and ToolCall.

Task completion is explicit: tool directives are the default, while a declared no-tool policy can finish or pause on a normal assistant response. No implicit "last message = answer" heuristic is used.

Repository Setup

cp local_settings.example.py local_settings.py
# Fill in your API keys and endpoints in local_settings.py

uv sync --dev
uv run pytest

Quick Start

CLI

uv run vv-agent --prompt "Summarize this framework" --backend moonshot --model kimi-k3

# With per-cycle logging
uv run vv-agent --prompt "Summarize this framework" --backend moonshot --model kimi-k3 --verbose

CLI flags: --settings-file, --backend, --model, --verbose.

Programmatic SDK

from vv_agent import Agent, RunConfig, Runner, function_tool

@function_tool
def read_order(order_id: str) -> str:
    """Read order information."""
    return "order details"

agent = Agent(
    name="ops",
    instructions="Check facts first, then answer.",
    model="kimi-k3",
    tools=[read_order],
)

result = Runner.run_sync(agent, "Analyze order 123", run_config=RunConfig(
    default_backend="moonshot",
))
print(result.status, result.final_output)

For defaults shared by several runs, create a configured Runner instead of repeating the same RunConfig:

runner = Runner.configured(RunConfig(
    model_provider=provider,
    model="kimi-k3",
    workspace="./workspace",
))
result = runner.run_sync(agent, "Analyze order 123")

Provider resolution is per-run then Runner. Model resolution is per-run, Agent, Runner, then the selected provider default. Model settings merge in the opposite layering direction: provider, Runner, Agent, then per-run, with each later layer overriding earlier fields.

Agent.output_type can coerce JSON final output into dict, list, dataclasses, or Pydantic-style models. Decorated tools may accept a leading ToolContext parameter; it is passed at invocation time and omitted from the tool JSON schema.

Streaming And Sessions

Ordinary Runner runs use a fresh SQLite :memory: store. Runner.start() returns a live RunHandle; events(), result(), cancel() and approve() share the same kernel path. Durable events derive from the session log; assistant deltas are volatile live observations. A JsonlRunEventStore is an optional projection sink.

from vv_agent import Agent, RunConfig, Runner

agent = Agent("assistant", "Answer briefly.", model="kimi-k3")
handle = Runner.start(agent, "Inspect the project", run_config=RunConfig(default_backend="moonshot"))
for event in handle.events():
    if event.type == "assistant_delta":
        print(event.delta, end="")
print(handle.result().final_output)

Use InteractiveAgentClient for multiple turns and SessionStore for durable retention. Waiting operations retain the same session_id and turn_id. After answering a parked approval, explicitly resume the handle or retained turn.

App Server

Use the App Server when a desktop app, worker, IDE, or other host process needs to drive vv-agent through a stable protocol instead of embedding the Python SDK directly. It runs JSONL over stdio, exposes Thread / Turn / Item lifecycle events, routes tool approval as server-to-client requests, supports thread/read and thread/resume replay, and exports typed JSON Schema and self-contained TypeScript bindings.

uv run vv-agent app-server --listen stdio --settings local_settings.py --backend moonshot --model kimi-k3
uv run vv-agent app-server schema --out ./app-server-schema
uv run vv-agent app-server generate-ts --out ./app-server-schema/typescript
uv run vv-agent debug app-server send-message "hello"

Product hosts implement AppServerHost to map product profiles, workspace context, tools, approval UI, memory, and model settings into framework Agent and RunConfig objects. The App Server remains a runtime boundary; it does not import product UI, account, billing, browser, or IM modules.

See docs/app-server.md for protocol details and docs/app-server-host-integration.md for the current host boundary and rollout checks.

Interactive Sessions

InteractiveAgentClient owns conversation turns, steering, follow-up, cancellation and approvals. Its default store is SQLite :memory:. Supply SQLiteStore or PostgresStore through AgentSessionOptions.session_store for durable retention. Creation-time session is a closed seed with messages and shared_state; messages and shared_state become read-only projections after creation.

from pathlib import Path
from vv_agent import Agent, AgentSessionOptions, InteractiveAgentClient, SQLiteStore, VvLlmModelProvider

with SQLiteStore.standalone("sessions.sqlite3") as store:
    if not store.connection.execute("PRAGMA user_version").fetchone()[0]:
        store.install_schema()
    client = InteractiveAgentClient(options=AgentSessionOptions(
        model_provider=VvLlmModelProvider(Path("local_settings.py"), default_backend="moonshot"),
        session_store=store,
    ))
    try:
        session = client.create_session(
            agent=Agent("assistant", "Remember prior turns.", model="kimi-k3"),
            session_id="thread-001",
        )
        print(session.prompt("Remember the project codename River.").final_output)
        print(session.prompt("What is the codename?").final_output)
    finally:
        client.driver.close()

Runner.resume(session_id, turn_id), AgentSession.continue_run() and App Server turn/resume drive an explicitly retained turn. Replies preserve turn budgets; fresh turns reset their counters. Closed sessions reject execution.

Agent As Tool, Handoff, And Policy

Use agent.as_tool() when a child agent should return a result to the parent agent and let the parent continue. Use handoff() when control should transfer to the target agent and the target output should finish the run.

from vv_agent import Agent, RunConfig, Runner, ToolPolicy, handoff
from vv_agent.constants import TASK_FINISH_TOOL_NAME

researcher = Agent(name="researcher", instructions="Collect facts.", model="kimi-k3")
writer = Agent(
    name="writer",
    instructions="Write from research.",
    model="kimi-k3",
    tools=[researcher.as_tool(name="research", description="Collect facts.")],
)
triage = Agent(
    name="triage",
    instructions="Transfer writing tasks.",
    model="kimi-k3",
    handoffs=[handoff(agent=writer, description="Use for writing.")],
)

result = Runner.run_sync(
    triage,
    "Write a short report.",
    run_config=RunConfig(
        default_backend="moonshot",
        max_handoffs=4,
        tool_policy=ToolPolicy(allowed_tools=[TASK_FINISH_TOOL_NAME, "transfer_to_writer"]),
    ),
)

A handoff is an outer Runner control transfer, not an agent-as-tool call. The target Agent resolves its own model and model settings, while the active session, cancellation token, and mutated shared state continue across the transition. max_handoffs defaults to 10 and limits control transfers independently from max_cycles. Approval resume preserves the same behavior.

No-tool completion is an explicit control, not a task or answer classifier. Set Agent(no_tool_policy="finish") when a normal assistant response should finish the run without task_finish, or override it for one call with RunConfig(no_tool_policy="continue" | "wait_user" | "finish"). Per-run configuration wins over a configured Runner default, which wins over the Agent value; omitting every layer uses finish. Inspect result.completion_reason, result.completion_tool_name, and result.partial_output to distinguish natural completion, tool-driven completion, waits, cancellation, failure, and max-cycle exhaustion.

RunConfig.budget_limits can independently limit total tokens, uncached input tokens, total or exact-name tool calls, active wall time, and host-metered cost. Limits are optional and task-neutral. Inspect result.budget_usage and result.budget_exhaustion; a budget stop is a typed failed result, not a successful answer. See Run Budgets.

Tools can request approval with @function_tool(needs_approval=True). By default the run enters WAIT_USER before the tool body is called and emits a ApprovalRequestedEvent. ToolPolicy(approval="never") disables that approval gate for trusted runs. The four policy modes are default, always, never, and on_request: default inherits the next configured policy, whereas explicit on_request follows each tool's static or dynamic approval declaration.

Custom tools may also attach an optional host-visible capability declaration:

from vv_agent import (
    ToolIdempotency,
    ToolMetadata,
    ToolPolicy,
    ToolSideEffect,
    function_tool,
)

@function_tool(
    tool_metadata=ToolMetadata(
        side_effect=ToolSideEffect.EXTERNAL,
        idempotency=ToolIdempotency.UNSUPPORTED,
        terminal=False,
        capability_tags=["ticket.write"],
        cost_dimensions=["support_api.request"],
    )
)
def create_ticket(title: str) -> dict[str, str]:
    return {"ticket_id": "TCK-1001", "title": title}

policy = ToolPolicy(
    denied_side_effects=[ToolSideEffect.EXECUTE],
    denied_capability_tags=["filesystem.delete"],
    deny_terminal_tools=True,
    denied_cost_dimensions=["gpu.second"],
)

side_effect is one coarse declaration with no inferred hierarchy; capability_tags and cost_dimensions are opaque exact-match labels, and cost dimensions are not measurements or prices. terminal=True only declares that a tool may return finish or wait_user; it never ends a run by itself. The four new policy fields are cumulative denials across Agent, configured Runner, per-run, and delegated-child layers, and a matching denial returns tool_not_allowed. They cannot grant a capability or remove an existing name, argument, approval, budget, or runtime restriction.

Typed metadata is separate from generic FunctionTool.metadata and is not added to the model-visible function schema. ToolMetadata.idempotency is the only idempotency declaration used by execution, telemetry, and retained operations.

Guardrails And Tracing

Input guardrails run before the model provider is called. Output guardrails run after a final output is available. Trace processors receive lightweight run and tool spans.

from vv_agent import Agent, GuardrailResult, RunConfig, Runner, input_guardrail

@input_guardrail
def reject_empty(ctx, input_text: str) -> GuardrailResult:
    del ctx
    if not input_text.strip():
        return GuardrailResult.block("input is required")
    return GuardrailResult.allow()

agent = Agent(
    name="assistant",
    instructions="Answer carefully.",
    model="kimi-k3",
    input_guardrails=[reject_empty],
)

result = Runner.run_sync(
    agent,
    "Summarize this project.",
    run_config=RunConfig(default_backend="moonshot", tracing={"workflow_name": "summary"}),
)

Shell Runtime Configuration (Windows)

bash runtime defaults are a startup/session configuration, not tool-call arguments.

• Run defaults: pass bash_shell, windows_shell_priority, and bash_env through RunConfig.metadata.
• Per-agent defaults: put the same keys in Agent.metadata.
• Recommended Windows priority: ["git-bash", "powershell", "cmd"]
• On Windows, bash-tool child processes default PYTHONUTF8=1 and PYTHONIOENCODING=utf-8 unless already overridden via the parent environment or bash_env.
• On Windows, bash-tool child processes are launched with hidden-console flags so GUI hosts can run bash / powershell commands without flashing a terminal window.
• Runner.run_sync(...) and Runner.stream_sync(...) both inherit compiled shell metadata.
• The bash tool schema description includes a runtime shell hint (resolved shell kind + invocation prefix), so the model sees which shell command style is expected before calling the tool.
• The runtime shell hint is frozen per task/session-run for local LLM requests to keep those request schemas stable across cycles and preserve prompt-cache efficiency. Frozen turn definitions retain the canonical schemas planned for the compiled task, so host-specific hint text does not affect the task-scoped toolset digest.
• Runner/CLI-generated tasks carry one resolved PromptBundle explicitly through AgentTask, each LlmRequest, the frozen turn definition and logged model operations. Generic metadata is not a prompt-section transport. Anthropic projection may use the canonical sections for cache breakpoints; other providers receive the deterministic flattened prompt.

from vv_agent import Agent, RunConfig, Runner

agent = Agent(
    name="desktop",
    instructions="Desktop helper",
    model="kimi-k3",
    metadata={"bash_env": {"HTTP_PROXY": "http://127.0.0.1:7890"}},
)
result = Runner.run_sync(
    agent,
    "Check the workspace.",
    run_config=RunConfig(
        default_backend="moonshot",
        metadata={
            "windows_shell_priority": ["git-bash", "powershell", "cmd"],
            "bash_env": {"PIP_INDEX_URL": "https://pypi.tuna.tsinghua.edu.cn/simple"},
        },
    ),
)

Session Execution

All public entrypoints use SessionDriver and the session kernel. SessionStore transactions own the log, inbox, leases, child admission and retained receipts. Ordinary runs use SQLite :memory:; durable SQLite/PostgreSQL are opt-in through the session API. Runner.start() provides non-blocking execution and cancellation.

from vv_agent import Agent, RunConfig, Runner

handle = Runner.start(Agent("assistant", "Answer briefly.", model="kimi-k3"), "Explain session history.",
                      run_config=RunConfig(default_backend="moonshot"))
# A host can call handle.cancel() from its UI or a timer.
print(handle.result().final_output)

See runtime-control.md for waits, explicit resume, LeaseLost backoff, budgets and event projections.

Runtime Log Payloads

The tool_result diagnostic contains the model-visible content, ordinary metadata, and a bounded content_preview; it does not duplicate artifact or cursor fields. Structured recovery belongs to ToolExecutionResult and is preserved in cycle results and retained operation receipts. A bounded bash result points to an immutable workspace artifact, while a bounded read_file result points to a source-verified cursor. Hosts must read artifacts through normal workspace policy; cursors reject changed sources, path mismatches, and invalid offsets.

Workspace Backends

Workspace file I/O is delegated to a pluggable WorkspaceBackend protocol. All built-in file tools (read_file, write_file, find_files, etc.) go through this abstraction.

find_files includes built-in safety defaults for large workspaces:

• Returns at most 100 paths per call by default (max_results can tune this, with hard cap).
• Uses ripgrep (rg) for fast local traversal when available, with automatic fallback to Python walk.
• search_files also uses rg for local workspaces (with Python fallback), defaults to smart-case matching (lowercase patterns are case-insensitive; patterns with uppercase stay case-sensitive), and skips hidden/common dependency roots unless explicitly included.
• search_files returns model-facing search text in ToolExecutionResult.content, while structured files/matches/counts live in ToolExecutionResult.metadata.
• Sensitive files such as .env and private keys are omitted by default; set include_sensitive=true to opt in.
• When listing from workspace root, common dependency/cache roots (for example node_modules, .venv, .git) are summarized instead of expanded.
• You can still inspect those paths explicitly by setting path to that directory (or by setting include_ignored=true).
• Supports scan_limit to stop early on very large trees; when triggered, response sets count_is_estimate=true.

from pathlib import Path
from vv_agent import RunConfig
from vv_agent.workspace import LocalWorkspaceBackend, MemoryWorkspaceBackend

local_config = RunConfig(workspace_backend=LocalWorkspaceBackend(Path("./workspace")))
memory_config = RunConfig(workspace_backend=MemoryWorkspaceBackend())

S3WorkspaceBackend

Install the optional S3 dependency: uv pip install 'vv-agent[s3]'.

from vv_agent.workspace import S3WorkspaceBackend

backend = S3WorkspaceBackend(
    bucket="my-bucket",
    prefix="agent-workspace",
    endpoint_url="https://oss-cn-hangzhou.aliyuncs.com",  # or None for AWS
    aws_access_key_id="...",
    aws_secret_access_key="...",
    addressing_style="virtual",  # "path" for MinIO
)

Custom Backend

Implement the WorkspaceBackend protocol declared in src/vv_agent/workspace/base.py to plug in any storage backend. A custom backend provides file enumeration, text/binary reads, writes, metadata, existence checks, file checks, and directory creation.

from vv_agent.workspace import WorkspaceBackend

class MyBackend(WorkspaceBackend):
    ...

Modules

Runtime Boundary

vv-agent owns the portable agent runtime: prompt assembly, model calls, tool planning, tool execution, memory compaction, typed events, cancellation, approval interruption, and replayable run history. Host products own product UI, user and workspace resolution, product storage, browser or IM integration, and the product-specific tools exposed to the model.

Host products should implement providers instead of patching vv-agent internals:

• AppServerHost maps product profiles, workspaces, tools, approval UI, memory, context, and model settings into App Server Agent and RunConfig objects when the host uses JSONL process integration.
• ApprovalProvider decides whether a tool call needs approval and returns the allow, deny, session-allow, or timeout decision from product UI or rules.
• ContextProvider contributes product prompt fragments such as profile, workspace, policy, or feature context before each run is compiled.
• vv_agent.memory.MemoryProvider connects product memory stores to memory search/save hooks and compaction lifecycle events.
• vv_agent.tools.ToolExecutor exposes product tools with schema, approval, timeout, error, and execution behavior. FunctionTool and @function_tool cover normal Python functions; custom executors are routed by ToolOrchestrator.
• SessionRunEventStore projects typed RunEvent history so app views can replay completed runs and parent/child run graphs.

This boundary keeps Agent, Runner, RunConfig, RunHandle, and RunEvent stable while allowing each host to keep its own account model, workspace model, storage backend, and UI workflow outside the framework.

Memory Compaction

MemoryManager measures context size in tokens and compacts history when the resolved auto-compaction threshold is exceeded.

• Task-level knobs:
  • memory_compact_threshold (default 250000; configured ceiling for full compaction)
  • memory_threshold_percentage (warning threshold percentage, default 90)
• Compile mapping:
  • AgentCompiler forwards stable agent/run metadata into AgentTask.
  • Resolved model limits are recorded as model_context_window and model_max_output_tokens; output capability is not copied into reserved_output_tokens.
  • Current durable turn definitions carry the exact configured threshold and capacity metadata used by resume.
  • Runtime-only compaction knobs remain metadata-backed until promoted into stable public fields.
• Token budget model:
  • Context precedence is explicit model_context_window, resolved model capability, then a derived planning context. The default is 250000 + 16000 + 13000 = 279000.
  • Output reserve precedence is effective ModelSettings.max_tokens, explicit reserved_output_tokens, then the 16000 framework fallback.
  • Only the framework fallback reserve may be capped downward by a smaller model_max_output_tokens; capability never overrides an explicit request or host reserve.
  • derived_prompt_capacity = max(model_context_window - reserved_output_tokens - autocompact_buffer_tokens, 0)
  • autocompact_threshold = min(memory_compact_threshold, derived_prompt_capacity); a configured threshold of zero selects the derived capacity, and a known derived capacity of zero stays zero.
  • The default autocompact buffer is 13000. MicrocompactionPolicy defaults to trigger/target ratios of 0.75/0.60, keeps 3 recent cycles, and only considers results longer than 500 characters.
• Effective-length strategy (backend-aligned):
  • If previous cycle token usage exists:
    • effective_length = previous_prompt_tokens + token_count(recent_tool_messages)
  • Otherwise fallback to:
    • vv_llm.chat_clients.utils.get_message_token_counts(...)
    • If tokenizer resolution fails, use a local CJK-aware estimate
• Compaction pipeline:
  1. One microcompaction pass archives eligible old tool results. Age is relative to assistant turns in the current transcript. Failed persistence keeps the original result; calls and results stay paired. A planned summary protects its raw tail.
  2. If still over threshold, summarize the previous summary and complete history prefix. memory_keep_recent_messages defaults to 10 raw messages; a cut inside a tool block moves left to retain the assistant and every corresponding result.
  3. Extract and normalize the model response, then replace the prefix only after effective-content, context-budget, recovery and token-reduction checks pass. Failure retains history; system messages and the raw tail remain unchanged.
  4. Summary metadata preserves artifact/cursor evidence deterministically. The model sees paths and retrieval hints; file references never trigger automatic reads.
  5. Prompt-too-long retries force a summary, then re-summarize with smaller tails. No successful shrink reaches the existing exhaustion boundary. Enabled Session Memory updates its baseline to the accepted summary plus tail.
• Compaction events:
  • New memory_compact_started producers include the typed trigger and the complete resolved capacity snapshot plus the micro target, candidate count, and estimated reclaimable tokens.
  • New memory_compact_completed producers include the strongest actual mode (none, micro, structural, summary, or emergency) and a content-aware changed flag plus archive count, actual reclaimed tokens, and artifact failure count.
  • Every current event includes the complete typed capacity and result fields; missing or unknown fields are rejected.
• Archive recovery:
  • Every tool defaults to ToolResultRetention.ARCHIVE; PRESERVE excludes a result only from proactive microcompaction.
  • Complete text is persisted through the effective WorkspaceBackend under the immutable logical .vv-agent/artifacts/ namespace before replacement. A failed or short write leaves the original message inline.
  • Microcompaction is disabled when model-visible read_file is unavailable, including use_workspace=False and explicit tool exclusion.
  • Existing typed artifacts are reused. The model sees only the compact marker's tool name, artifact path, retrieval hint, and bounded excerpt.
• Configure proactive compaction with RunConfig(microcompaction_policy=MicrocompactionPolicy(...)). The policy is copied to AgentTask and frozen/restored under runtime_controls.microcompaction_policy in the run definition.
• The model-visible replacement has this closed shape; byte size and SHA-256 remain only in the host-visible ToolArtifactRef:

<Tool Result Compact>
tool_name: web_search
artifact_path: .vv-agent/artifacts/<run>/<call>.txt
retrieval_hint: use read_file on artifact_path if needed
excerpt:
<bounded head/tail preview>
</Tool Result Compact>

• Session Memory behavior:
  • Stored in workspace/.memory/session/<session-or-task-scope>/session_memory.json by default
  • Scoped to the current session when metadata.session_id is present; otherwise scoped to the current task_id
  • New sessions/tasks start without inherited Session Memory from previous sessions/tasks
  • Loaded once when a new run is compiled and frozen into its first system message as <Session Memory>; every cycle reuses the same PromptBundle
  • Entries extracted during the active run are persisted but become visible only when the next new run is compiled
  • Retained-turn resume reuses the frozen memory section without rereading the store or rewriting the active prompt
  • Extraction reuses the configured memory summary backend/model
  • Full compaction resets transcript tracking but preserves persisted memory entries
  • Sub-tasks disable Session Memory by default to avoid parent/child memory-file contamination

Runtime metadata keys

Pass these via Agent.metadata or RunConfig.metadata; the compiler forwards them into AgentTask.metadata:

• memory_keep_recent_messages
• model_context_window
• model_max_output_tokens (resolved model capability; not an implicit request limit)
• reserved_output_tokens
• autocompact_buffer_tokens
• include_memory_warning
• session_memory_enabled
• session_memory_min_tokens
• session_memory_max_tokens
• session_memory_min_text_messages
• session_memory_storage_dir
• tool_result_excerpt_head
• tool_result_excerpt_tail
• summary_event_limit

Memory summary model selection priority

Priority is strict:

1. AgentTask.metadata.memory_summary_model, with optional memory_summary_backend.
2. The current task model through the run's ModelProvider.

Built-in Tools

find_files, file_info, read_file, write_file, edit_file, search_files, todo_write, ask_user, bash, check_background_command, stop_background_command, read_image, create_sub_task, sub_task_status, activate_skill.

Custom tools can be registered via ToolRegistry.register().

bash waits up to yield_time_ms (default 1000, integer 0..10000) for the command to finish, then returns its current output and a session_id while the process keeps running. Zero requests an immediate handle. Optional timeout_seconds (integer 1..86400) sets one execution deadline from process start; omission means no execution deadline. Querying does not extend it.

check_background_command({"session_id":"bg_..."}) reads a current snapshot; stop_background_command({"session_id":"bg_..."}) requests process-tree termination. A successful start or running query is a completed SUCCESS / continue management receipt, so the session kernel can call its model again. The actual process state is in content and metadata. Nonzero observed exit codes remain errors. Sessions belong to their initiating task and workspace; a missing local record does not establish whether an external process exited. Live oversized output includes an immutable artifact for complete recovery. See Bash process management for examples and termination observations.

Sub-agents

Use Agent.as_tool() when the parent agent should call a child agent and then continue. Use handoff() when the child agent should take over and finish the run. Use create_sub_task and sub_task_status when the model needs explicit background or parallel task management.

Each delegated task is an independently scheduled kernel session. Parent admission freezes its identity, prompt, policy, model and budget; the parent adopts only its authenticated terminal. Intermediate user waits stay on the child. Batch children can drive independently after the parent releases its lease.

Use sub_task_status to read owner-scoped progress, wait for completion or admit a continuation message. Hosts steer a retained child by pushing a targeted steer inbox item through its store or interactive session. Child lifecycle events carry parent and child identities for subscription and replay.

Configured child runs inherit the same explicit ModelProvider as the parent and resolve their own model. No settings path or backend fallback is rebuilt inside the child runtime.

Examples

The examples/ directory now contains public SDK cookbook scripts plus a small set of lower-level runtime integration examples. See examples/README.md for the full list.

uv run python examples/01_quick_start.py
uv run python examples/24_workspace_backends.py

Testing

uv run pytest                              # unit tests (no network)
uv run ruff check .                        # lint
uv run ty check                            # type check

VV_AGENT_RUN_LIVE_TESTS=1 uv run pytest -m live   # integration tests (needs real LLM)

Environment variables for live tests: