- Home
- Skills
- Agents & Orchestration
- FPGA 03/13 · Testbench Claude — Self-Checking Tests
Works with the AI Agents you already use
FPGA 03/13 · Testbench Claude — Self-Checking Tests
Builds machine-verifiable, file-driven HDL testbenches with independent golden references and negative controls.
$10
FPGA 03/13 · Testbench Claude — Self-Checking Tests
Example session with this skill installed
Generate a self-checking Verilog testbench for a 16-bit pipelined multiplier. Use external hex files for vectors and include a monitor task that counts errors against a golden model.
- Read your context and instructions
- Compiled the fpga 03 13
I have generated tb_multiplier.v. It features decoupled stimulus and monitor tasks, uses $readmemh to load tb/multiplier_vectors.hex, and performs comparisons at the end of the time step. A negative control pass verified the monitor by injecting a bit-flip in the DUT instantiation.
Connects securely to your tools. The creator never sees your data.
About this skill
The problem
Manual waveform inspection is slow and misses silent errors like single-cycle timing shifts. Without independent reference models, developers often build testbenches that share the same logic bugs as their RTL, leading to false passes that fail after synthesis.
What it does
- Generates self-checking HDL testbenches with machine-readable PASS/FAIL verdicts.
- Separates clock, reset, stimulus, and monitor tasks into independent processes to prevent race conditions.
- Implements file-driven verification using external .hex and .json vectors for reconfigurability without recompilation.
- Executes negative control tests against deliberately broken DUT variants to validate the checker's effectiveness.
- Applies biased random generation and exhaustive testing for small bit-widths to hit difficult corner cases.
Frameworks & tools
Verilog, SystemVerilog, HDL simulators (Icarus, Vivado, ModelSim), Python-based golden models.
Why this beats prompting it yourself
Standard LLM prompts often produce "first-fail" testbenches that lack proper timing synchronization or independent oracles. This skill enforces a strict verification architecture, including delta-cycle settled comparisons and mandatory negative controls, ensuring your test suite actually catches bugs instead of just printing signals.
Use cases
- Verifying complex RTL modules against a Python-based golden reference.
- Building regression suites that load test vectors from external files.
- Stress-testing arithmetic units with biased random stimulus for boundary values.
- Proving testbench reliability via fault-injection and negative control runs.
Known limitations
Requires an existing golden model or oracle definition to establish independence. Optimized for Verilog/SystemVerilog workflows.
Requires an independent golden reference (like a Python model) to function. Specifically designed for Verilog/SystemVerilog simulation environments.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 2 months ago