WorkerLane

Harness-agnostic agent coworkers for any agent runtime. One package that wires team agents, approvals, memory, traces, and handoff contracts into whatever harness you already use.

Part of Talocode — open tools people trust, hosted power behind them.

What it is

AI agents are great at running tasks. The hard part is everything around them: who can run what, what gets approved, what happened last session, and proving it after the fact. WorkerLane bundles those pieces into one package that works with any agent harness — no runtime brand loyalty, no side to pick.

• AI coworkers — create agents with a name and role, queue tasks, gate destructive work behind approvals
• Agent-human handoffs — validate and repair step-to-step payloads against schema contracts so runs never silently drift
• Full data recording, owned by you — trace spans with cost and status, plus durable memory that persists across sessions. Everything is stored locally in ~/.workerlane unless you opt into the hosted path
• Computer use — pair with @talocode/screenlane (screenlane mcp) to add screen capture, dictate, and command tools to the same harness
• Works with any harness — expose the whole bundle as MCP tools and connect any MCP-compatible agent

Why it exists

The catalog is strong but fragmented — agents, traces, memory, handoffs each ship separately. WorkerLane removes the assembly. One install, one surface, all the pieces wired together. It is positioned as a capability layer above the harness, exactly where compounding value lives.

Install

pip install workerlane

or

npm install -g @talocode/workerlane

No runtime, no cloud required for the local engine. The Python SDK uses a globally installed workerlane engine when available, otherwise it runs the public npm package through npx. State lives in ~/.workerlane/ (override with WORKERLANE_DIR).

Quickstart (Python SDK)

from workerlane import WorkerLaneClient

lane = WorkerLaneClient()

# create a coworker
agent = lane.create_agent(name="qa-bot", role="reviews every PR before merge")

# queue a task, gated behind a human approval
run = lane.run_agent(agent["id"], task="audit the auth flow", require_approval=True)
# run["status"] == "pending-approval"

# a human approves it — the decision is recorded
approved = lane.approve_run(run["id"], True)

# hosted path (needs TALOCODE_API_KEY)
hosted = lane.hosted()
hosted.agent_run(agent["id"], task="audit the auth flow")

Auth / env

• TALOCODE_API_KEY — set to use the hosted /v1/workerlane/* path (Stacklane billing)
• TALOCODE_BASE_URL — defaults to https://api.talocode.site
• WORKERLANE_DIR — where local state is stored (defaults to ~/.workerlane)

API surface

The Python SDK mirrors the CLI and the hosted API:

Full MCP surface (workerlane mcp) exposes 16 tools: workerlane_agent_create/list/run/approve/complete/runs, workerlane_handoff_validate/repair, workerlane_trace_start/span/span_end/complete/list, workerlane_memory_remember/recall/list.

CLI (npm engine)

workerlane agent create --name "qa-bot" --role "reviews every PR"
workerlane agent run --agent <ID> --task "audit" --approve
workerlane agent approve --run <RUN_ID> --yes
workerlane mcp          # start the MCP server for any harness
workerlane pricing
workerlane capabilities

Related packages

• @talocode/workerlane (npm engine) · pip install workerlane (this package)
• @talocode/screenlane · pip install talocode-screenlane
• @talocode/worklane · pip install talocode-worklane
• @talocode/memorylane · pip install talocode-memorylane
• @talocode/handofflane
• @talocode/tracelane

Talocode ecosystem

More: github.com/talocode · talocode.site · docs.talocode.site

License

MIT © Talocode