ScreenLane (Python)
Screen-aware voice / text command layer for AI agents — capture screen context, dictate intent, route commands to Codra, Tera, and other targets.
Part of Talocode.
| Package | pip install talocode-screenlane |
| Import | from talocode_screenlane import ScreenLaneClient |
| CLI | screenlane-py |
| Local default | http://127.0.0.1:3070 |
| Cloud | https://api.talocode.site + TALOCODE_API_KEY |
| Repo | github.com/talocode/screenlane |
| License | MIT |
Why ScreenLane?
Agents often lack what’s on the user’s screen. ScreenLane sits between:
- Capture — screen / context text
- Intent — voice transcript or typed command
- Route — structured command for Codra, Tera, GateLane, etc.
Open-source local server + cloud routes so you can demo offline and scale with Talocode Cloud.
Install
pip install -U talocode-screenlane
Python 3.10+.
Quickstart
Local server
# In the ScreenLane Node package / monorepo:
# screenlane serve → http://127.0.0.1:3070
from talocode_screenlane import ScreenLaneClient
client = ScreenLaneClient() # local default
print(client.health())
print(client.doctor())
print(client.demo())
result = client.command({
"text": "Fix this error",
"contextText": "TypeError: Cannot read properties of undefined",
"target": "codra",
})
print(result)
Android workflow foundation
ScreenLane plans Android workflows locally before a future companion bridge can execute anything. Inspection can be planned without approval. Tap, type, and scroll require explicit approval. Sensitive app classes are blocked, and every request records a local receipt.
from talocode_screenlane import ScreenLaneClient
client = ScreenLaneClient()
print(client.android_doctor())
pending = client.plan_android_action({
"action": "tap",
"appId": "com.example.notes",
"target": "New note",
})
print(pending["status"]) # pending_approval
approved = client.plan_android_action({
"action": "tap",
"appId": "com.example.notes",
"target": "New note",
"approved": True,
})
print(approved["status"])
print(client.list_android_receipts())
The local API exposes GET /v1/screenlane/android/doctor, POST /v1/screenlane/android/actions, and GET /v1/screenlane/android/receipts.
Cloud
import os
from talocode_screenlane import ScreenLaneClient
cloud = ScreenLaneClient(
base_url="https://api.talocode.site",
api_key=os.environ["TALOCODE_API_KEY"],
)
cloud.command({
"text": "Summarize this screen",
"contextText": "...",
"target": "tera",
})
Environment
| Variable | Purpose |
|---|---|
TALOCODE_API_KEY |
Cloud auth |
TALOCODE_BASE_URL |
Optional cloud base |
| Local URL | Passed as base_url / client default http://127.0.0.1:3070 |
Client API
| Method | Purpose |
|---|---|
health() |
Liveness |
doctor() |
Dependency / config checks |
android_doctor() |
Android companion readiness and safety policy |
plan_android_action(payload) |
Plan an Android inspection or approval-gated action |
list_android_receipts() |
List local Android workflow receipts |
capture(payload) |
Screen / context capture |
dictate(payload) |
Speech / text dictation |
command(payload) |
Build / run agent command |
send(payload) |
Dispatch to target |
list_contexts() |
Known contexts |
list_commands() |
Command catalog |
demo() |
Deterministic demo flow |
Typical payload fields: text, contextText, target, sessionId, offline flags (see CLI).
CLI
screenlane-py --help
screenlane-py demo
screenlane-py command --text "Fix this error" --context-text "..." --offline
How it fits Talocode
User screen / voice
↓
ScreenLane (this)
↓
Codra / Tera / GateLane / custom target
Pair with Codra for code fixes and Tera for general LLM actions.
Related packages
| Package | Role |
|---|---|
talocode-screenlane |
This package |
talocode-codra |
Coding agent target |
talocode-tera |
Chat / rewrite target |
talocode |
Unified cloud client |
Talocode ecosystem
| Project | Role |
|---|---|
| ScreenLane | This product |
| Codra | Coding agent |
| Tera | AI capabilities |
| GateLane | MCP gateway |
| StackLane | Cloud control plane |
| SearchLane | Search capabilities for agents |
| Agent Browser | Browser automation |
| ClipLoop | Short-form video workflows |
| InvoiceLane | Invoicing workflows |
| ContextLane | Context ingestion |
| MemoryLane | Agent memory |
| Tradia | Trading intelligence |
More: github.com/talocode · talocode.site · docs.talocode.site
License
MIT © Talocode
Metadata
Release files for talocode-screenlane 0.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| talocode_screenlane-0.2.0.tar.gz | 6.7 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| talocode_screenlane-0.2.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 13.6 kB
Release files / talocode_screenlane-0.2.0.tar.gz
| Download URL | talocode_screenlane-0.2.0.tar.gz |
|---|---|
| Size | 6.7 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
3ad38c894fa68608ad1bd96495e81a8283622c1c5b15500bfc2ff1eea646f056
|
|
BLAKE2b-256 checksum How to use checksums |
09db8d8b92dfdb3a979c74cd5fd8e71c408d60aadd056181e19a4d4748a6f278
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.5
|
Release files / talocode_screenlane-0.2.0-py3-none-any.whl
| Download URL | talocode_screenlane-0.2.0-py3-none-any.whl |
|---|---|
| Size | 6.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
1fa861d20318f56bd445d911a404f1b5fea1a4235e322a109a3cbff91472c6ed
|
|
BLAKE2b-256 checksum How to use checksums |
99230184c293b2ded5d796d22fadd12f52a3e5b15d29dac4f960a543c5698848
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.13.5
|