01 — CODE
The repositories, explained.
What each repo actually contains and where to start reading. Trees reflect the real structure, not an idealized one — key files marked.
GauthamCodes/coco-robot-ros2
Two ROS 2 packages behind the IEEE ICRM 2025 paper.
The simulation platform behind the IEEE ICRM 2025 paper.
gazebo_modelsholds the URDF authored from CAD meshes plus the worlds and launch files;custom_teleopdrives the base and the manipulator from a keyboard.- coco-robot-ros2/
- gazebo_models/# robot description + worlds
- urdf/# authored from CAD meshes
- meshes/
- launch/
- package.xml
- custom_teleop/# base + arm keyboard teleop
- launch/
- setup.py
- docs/images/
- README.md
GauthamCodes/PAAMB-Bioinspired-Robotic-Snake-for-Pipeline-Surveillance
The whole snake in one repo — firmware, dashboard, model, STLs.
Three firmware sketches (motor, sensor, camera), the Flask dashboard with its on-device FOMO/TFLite detector, the printable STLs — and
secrets.h.example, the artifact of the hardcoded-WiFi lesson from WHAT BROKE.- PAAMB-…-Surveillance/
- paamb_motor/# ESP32 + PCA9685 gait firmware
- paamb_sensor/# MQ135 + BMP280 telemetry
- paamb_cam/# ESP32-CAM stream
- dashboard_app.py# Flask + FOMO/TFLite overlay
- model.tflite# the deployed detector
- design/# printable STLs
- secrets.h.example# the hardcoded-WiFi lesson
- README.md
GauthamCodes/LeRobot-ST3215-Controller
Teach-and-replay over a serial-bus servo protocol.
The teach-and-replay stack for six daisy-chained ST3215 servos. The
docs/folder is the honest part: WIRING, SERVO_IDS, and TROUBLESHOOTING exist because the bus faults in the case study actually happened.- LeRobot-ST3215-Controller/
- scripts/
- teach_replay.py# 3 threads @ 50 Hz
- teach_mode.py
- realtime_feedback.py
- scan_ids.py# bus utilities
- scservo_sdk/# ST3215 protocol
- protocol_packet_handler.py
- group_sync_write.py
- docs/
- WIRING.md
- TROUBLESHOOTING.md# the bus faults, documented
- requirements.txt
GauthamCodes/ai-ap-employee
A supervisor/worker pipeline with safety gates before payment.
Each agent is its own module under
agents/, the safety gates live in the approval path, andtests/exercises the whole flow end to end.- ai-ap-employee/
- agents/# 6 specialized agents
- supervisor.py# the orchestrator
- ingestion.py
- matching.py
- accounting.py
- approval.py# the 3 safety gates
- schemas/state.py# shared pipeline state
- tools/connectors.py
- tests/# 7/7 end-to-end
- main.py# FastAPI entry