Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Power Profiler Skill · a5c-ai | DecimalAI

a5c-ai/power-profiler

SkillScore86Maintained

Power consumption measurement and analysis expertise for embedded systems. Integrates with power analyzer tools to measure, profile, and optimize power consumption in battery-powered and energy-efficient designs.

Testingby a5c-ai

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

QA & Test Engineering pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/a5c-ai-power-profiler/SKILL.md

decimalai skills pull a5c-ai-power-profiler --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/a5c-ai-power-profiler/SKILL.md https://app.decimal.ai/s/a5c-ai-power-profiler/SKILL.md

Raw file: /s/a5c-ai-power-profiler/SKILL.md ↗ · machine-readable summary at /s/a5c-ai-power-profiler.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 26d ago

Yes — +45% lift · at higher cost+51% tokens0% turns86% of whole cases pass with skill (19/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-01	✗→✓	▲ Improved	-13%	0%
case-03	✗→✓	▲ Improved	11%	0%
case-06	✗→✓	▲ Improved	75%	0%
case-08	✗→✓	▲ Improved	187%	0%
case-09	✗→✓	▲ Improved	52%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

a5c-ai / babysitter

library/specializations/embedded-systems/skills/power-profiler/SKILL.md

View on GitHub ↗

License MITImported 8/10/2026Version v1

Skill contents

SKILL.md

Power Profiler Skill

Expert skill for power consumption measurement, analysis, and optimization in embedded systems. Provides integration with power analyzer tools and deep expertise in low-power design techniques.

Overview

The Power Profiler skill enables comprehensive power analysis for embedded systems, supporting:

Power analyzer tool integration (Otii Arc, Nordic PPK2, Joulescope)
Current measurement configuration and calibration
Power mode transition analysis
Battery life estimation calculations
Power profile comparison and trending
Peripheral power contribution analysis
Sleep mode leakage identification
Energy-per-operation measurements

Capabilities

1. Power Measurement Configuration

Configure and calibrate power measurement hardware:

javascript
// Example: Configure Otii Arc power analyzer
const powerConfig = {
  analyzer: 'otii-arc',
  sampleRate: 4000,        // Hz
  currentRange: 'auto',     // auto, low (uA), high (mA)
  voltageOutput: 3.3,       // Supply voltage
  triggerMode: 'gpio',      // gpio, serial, manual
  triggerPin: 'GPI1'
};

2. Power State Analysis

Analyze power consumption across different operating states:

markdown
## Power State Profile

| State | Current (avg) | Duration | Energy |
|-------|---------------|----------|--------|
| Active | 15.2 mA | 50 ms | 760 uJ |
| Processing | 45.3 mA | 10 ms | 453 uJ |
| Radio TX | 120 mA | 5 ms | 600 uJ |
| Sleep | 2.5 uA | 935 ms | 2.3 uJ |
| **Total Cycle** | - | 1000 ms | **1815.3 uJ** |

Average Current: 1.815 mA
Battery Life (1000 mAh): 551 hours (23 days)

3. Power Mode Transition Analysis

Identify and analyze power state transitions:

Wake-up latency measurement
Sleep entry timing
Transition energy overhead
Unexpected wake-up detection
Power state sequence verification

4. Peripheral Power Contribution

Break down power consumption by peripheral:

markdown
## Peripheral Power Breakdown

| Peripheral | Active Current | Sleep Current | Contribution |
|------------|----------------|---------------|--------------|
| MCU Core | 8.5 mA | 1.2 uA | 35% |
| Radio (BLE) | 6.2 mA | 0.5 uA | 25% |
| Sensors | 3.8 mA | 0.8 uA | 16% |
| Display | 4.2 mA | 0.1 uA | 17% |
| Other | 1.5 mA | 0.4 uA | 7% |

5. Battery Life Estimation

Calculate expected battery life for various usage scenarios:

javascript
// Battery life estimation parameters
const batteryEstimate = {
  batteryCapacity: 230,     // mAh (CR2032)
  dutyCycle: {
    activePeriod: 100,      // ms
    sleepPeriod: 9900,      // ms
    transmitCount: 1        // per cycle
  },
  currentProfile: {
    active: 15.0,           // mA
    sleep: 2.5,             // uA
    transmit: 120.0         // mA
  },
  derating: 0.85            // 85% capacity utilization
};

// Calculated: 2.3 years battery life

Process Integration

This skill integrates with the following processes:

| Process | Integration Point | |---------|-------------------| | power-consumption-profiling.js | Primary execution - all phases | | low-power-design.js | Measurement and validation phases | | real-time-performance-validation.js | Power budget verification |

Tool Integration

Supported Power Analyzers

| Tool | Features | Connection | |------|----------|------------| | Otii Arc | High precision, automation API | USB, REST API | | Nordic PPK2 | Source/ampere meter modes | USB, nRF Connect | | Joulescope | Real-time streaming, triggers | USB, Python API | | Keysight N6705C | Multi-channel, high accuracy | GPIB, USB, LAN | | Qoitech Otii | Cloud integration, sharing | USB, Otii Desktop |

Data Export Formats

CSV time-series data
JSON power profiles
PNG/SVG visualizations
Interactive HTML reports
Otii project files (.otii)
Joulescope capture files (.jls)

Workflow

1. Setup Measurement Environment

bash
# Verify power analyzer connection
otii-cli device list

# Configure measurement parameters
otii-cli project create \
  --name "power-profile-$(date +%Y%m%d)" \
  --voltage 3.3 \
  --current-range auto

2. Capture Power Profile

bash
# Start recording with GPIO trigger
otii-cli recording start \
  --trigger gpio:GPI1:rising \
  --duration 10s

# Or use serial trigger
otii-cli recording start \
  --trigger serial:START \
  --stop-trigger serial:STOP

3. Analyze Results

bash
# Export measurement data
otii-cli recording export \
  --format csv \
  --output power-data.csv

# Generate statistics
otii-cli statistics \
  --markers state:active,state:sleep \
  --output stats.json

4. Generate Report

The skill generates comprehensive power analysis reports including:

Executive summary with key metrics
State-by-state power breakdown
Transition timing analysis
Battery life projections
Optimization recommendations
Comparison with targets/baselines

Output Schema

json
{
  "summary": {
    "averageCurrent_mA": 1.815,
    "peakCurrent_mA": 120.0,
    "sleepCurrent_uA": 2.5,
    "estimatedBatteryLife_hours": 551
  },
  "powerStates": [
    {
      "name": "active",
      "current_mA": 15.2,
      "duration_ms": 50,
      "energy_uJ": 760
    }
  ],
  "transitions": [
    {
      "from": "sleep",
      "to": "active",
      "latency_us": 125,
      "energy_uJ": 1.2
    }
  ],
  "peripheralBreakdown": {
    "mcu": { "active_mA": 8.5, "sleep_uA": 1.2 },
    "radio": { "active_mA": 6.2, "sleep_uA": 0.5 }
  },
  "recommendations": [
    "Reduce radio TX power by 3dB to save 15% energy",
    "Enable peripheral clock gating during sleep"
  ],
  "artifacts": [
    "power-profile.csv",
    "power-report.html",
    "waveform.png"
  ]
}

Best Practices

Measurement Setup

Use kelvin sense connections for accurate voltage measurement
Minimize wire length between analyzer and DUT
Ensure stable power supply to analyzer
Allow thermal stabilization before measurement

Calibration

Zero-offset calibration before each session
Verify measurement accuracy with known load
Document measurement uncertainty

Analysis

Use markers to identify power states
Compare against power budget requirements
Track power metrics across firmware versions
Document measurement conditions

References

Nordic PPK2 User Guide
Joulescope User Guide
Otii Arc Documentation
"Power Management for Internet of Things" - ARM
Low-Power Design Methodology Manual

MCP Server Integration

Compatible MCP servers for enhanced functionality:

| Server | Purpose | |--------|---------| | embedded-debugger-mcp | Coordinate debug probes with power measurement | | serial-mcp-server | Serial trigger synchronization | | tinymcp | Device state monitoring |

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 We have an Otii Arc connected to our custom BLE board. I need a CLI setup script to capture current consumption with a GPIO rising edge trigger for a 10-second window at 3.3V. After analyzing the capture, output a JSON profile containing summary statistics (average and peak current in mA, sleep current in uA, estimated battery life in hours), state-by-state energy breakdowns, transition latency/energy, peripheral power split, recommendations for lowering energy, and generated artifact filenames.	fail→pass	45,932	27,816	-39%	1	1	0%	7,899	6,886	-13%	0	0	—
case-02 We recorded trace data from a Nordic PPK2 during state transitions on our embedded micro-controller. Can you analyze the latency and energy overhead associated with moving between sleep mode and active processing, quantify individual peripheral contributions (MCU core, radio, sensors, display), and suggest low-power hardware and software optimizations to reduce sleep leakage?	fail→fail	31,004	28,454	-8%	1	1	0%	5,455	7,748	+42%	0	0	—
case-03 Write the CLI shell command sequence to initialize a project for current measurement at 3.3 volts with automatic current range selection.	fail→pass	29,244	3,722	-87%	1	1	0%	2,264	2,503	+11%	0	0	—
case-04 We are selecting hardware analyzers for our IoT testbench. We need a tool that supports both Source mode and Ampere meter mode over USB using nRF Connect, and another tool that provides high precision REST API automation. List the matching analyzers and their primary connection interfaces.	pass→pass	9,443	4,429	-53%	1	1	0%	1,698	2,747	+62%	0	0	—
case-05 Calculate the expected battery life in hours and years for an IoT sensor running on a coin cell with 230 mAh capacity. The device runs a cycle of 100 ms active (15.0 mA), 100 ms transmit (120.0 mA), and 9800 ms sleep (2.5 uA), with an 85% capacity utilization factor applied.	pass→pass	16,589	17,440	+5%	1	1	0%	2,933	6,009	+105%	0	0	—
case-06 Create a peripheral energy audit table for an nRF52-based tracker detailing MCU core (8.5 mA active, 1.2 uA sleep), BLE Radio (6.2 mA active, 0.5 uA sleep), and OLED Display (4.2 mA active, 0.1 uA sleep). Include percentage contributions.	fail→pass	13,983	14,880	+6%	1	1	0%	2,965	5,192	+75%	0	0	—
case-07 Provide a state-by-state power consumption table for an active-processing-sleep cycle, showing average current, state duration, and calculated energy per state.	fail→fail	21,138	16,496	-22%	1	1	0%	3,483	5,718	+64%	0	0	—
case-08 Construct a JSON summary object for a power profiling session recording 1.815 mA average current, 120.0 mA peak current, 2.5 uA sleep current, and 551 hours battery life.	fail→pass	5,521	3,350	-39%	1	1	0%	841	2,416	+187%	0	0	—
case-09 When analyzing micro-controller power state transitions from deep sleep to active execution, what key timing and energy metrics should be extracted from waveform markers?	fail→pass	17,661	19,061	+8%	1	1	0%	3,062	4,646	+52%	0	0	—
case-10 To avoid IR drop voltage measurement errors when sampling high dynamic range pulses (1 uA to 150 mA) on a target board, what physical wiring methodology must be used between the power profiler and the DUT?	pass→pass	13,923	11,661	-16%	1	1	0%	1,853	3,845	+108%	0	0	—
case-11 What calibration procedure should be executed on current measurement hardware prior to beginning a high-resolution power profiling session?	pass→pass	18,404	14,114	-23%	1	1	0%	2,531	4,251	+68%	0	0	—
case-12 Which Model Context Protocol (MCP) server integration should be utilized to synchronize hardware debug probes with dynamic current profiling measurements?	fail→pass	17,653	2,774	-84%	1	1	0%	2,788	2,376	-15%	0	0	—
case-13 Map the execution phases of low-power firmware development to their corresponding process scripts for power consumption profiling, low power design validation, and real-time power budget verification.	fail→fail	30,362	22,773	-25%	1	1	0%	4,742	5,459	+15%	0	0	—
case-14 List the standard native binary capture file formats supported for hardware-specific desktop profiling applications like Otii and Joulescope.	pass→pass	18,936	13,620	-28%	1	1	0%	2,518	3,883	+54%	0	0	—
case-15 Provide the CLI shell command to start an Otii profiling capture using UART string triggers 'START' and 'STOP'.	fail→pass	10,886	2,482	-77%	1	1	0%	1,598	2,272	+42%	0	0	—
case-16 In an embedded system where peripheral leakage accounts for 30% of sleep mode current, what software optimization recommendation should be generated in the audit report?	pass→pass	13,655	12,685	-7%	1	1	0%	1,813	3,857	+113%	0	0	—
case-17 Structure a JSON transitions array element representing a state shift from sleep to active with 125 us latency and 1.2 uJ energy cost.	fail→pass	6,041	3,464	-43%	1	1	0%	1,149	2,447	+113%	0	0	—
case-18 Write the command to extract statistical metrics from an Otii CLI recording using state markers 'active' and 'sleep' to produce a JSON report file.	fail→pass	8,659	3,372	-61%	1	1	0%	1,396	2,387	+71%	0	0	—
case-19 When estimating real-world battery operating life for lithium coin cells from calculated average load currents, what standard capacity utilization derating factor should be applied?	fail→pass	16,080	13,854	-14%	1	1	0%	2,599	4,355	+68%	0	0	—
case-20 We are designing the power distribution network (PDN) schematic for a 4-layer PCB running an STM32WB55. Select target impedance values and ceramic decoupling capacitor values (0.1uF vs 10uF) for filtering 2.4GHz RF ripple.	pass→pass	29,502	29,217	-1%	1	1	0%	4,710	7,793	+65%	0	0	—
case-21 Write a C code implementation of a priority-preemptive FreeRTOS task scheduler algorithm that handles priority inversion using mutex lock inheritance on Cortex-M4.	pass→pass	42,817	49,014	+14%	1	1	0%	6,879	10,201	+48%	0	0	—
case-22 Explain the chemical redox equations during discharge of Lithium Thionyl Chloride (LiSOCl2) primary cells compared to Lithium Manganese Dioxide (LiMnO2), focusing on passivating layer growth.	pass→pass	23,834	29,340	+23%	1	1	0%	3,711	7,203	+94%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +45 percentage points is the difference between those two pass rates over the 22 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 40.9%
Pass rate with the skill: 86.4%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 0
Skill version measured: v1
Run completed: 8/29/2026, 14:18 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-01

Prompt

We have an Otii Arc connected to our custom BLE board. I need a CLI setup script to capture current consumption with a GPIO rising edge trigger for a 10-second window at 3.3V. After analyzing the capture, output a JSON profile containing summary statistics (average and peak current in mA, sleep current in uA, estimated battery life in hours), state-by-state energy breakdowns, transition latency/energy, peripheral power split, recommendations for lowering energy, and generated artifact filenames.

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run86

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+74% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 26 days ago

Other measured skills in the registry, with their headline benchmark lift.