Techvify Forge Pulse · Performance

Know it scales before you ship.

Forge Pulse is Techvify’s performance-testing practice, built on an upstream open-source load generator. The agent bootstraps idiomatic load, stress, and soak tests; a tool bridge validates scripts and runs bounded checks from your editor; results stream live to a dashboard when you configure an output; and thresholds turn a perf regression into a failed run — caught before production, not after.

6
Bridge tools
50 VU
In-editor cap
p(95)
Threshold gate

What's inside

Describe the test. Ship the proof.

AI agent bootstrap

A subcommand extension installs skill bundles and registers the load-testing tool bridge into your editor. Describe a test in plain English and the agent writes an idiomatic script.

Tool bridge · 6 tools

Six tools cover it: validate a script, run a bounded check in-editor (capped at 50 VUs / 5 min), and look up documentation. Full load tests run from the CLI or CI.

Load, stress, and soak

Ramping-VU scenarios with thresholds. p(95) < 200 ms latency and < 1% errors are enforced per run — a breach fails the run.

Live dashboard streaming

Point the runner at a hosted metrics output and latency, throughput, and error rate stream live during the run — regressions surface mid-run, not in production.

Browser + assertions

Browser testing covers a growing subset of the standard automation API; a companion assertion library adds an expect() API — non-retrying for static values, auto-retrying for async UI.

A load test is a small script the agent writes for you: ramp to a target of virtual users, assert on status and latency, and declare thresholds that decide pass or fail. Version it beside the code it tests.

What a run looks like

Thresholds gate the run.

The runner fails the run when a threshold is breached, so a performance regression stops the pipeline the same way a failing unit test does. The same metrics stream live into your dashboards while the test runs.

$ run perf/load.js
passed: status is 200checks  100.00%  ✓ 14820  ✗ 0http_req_duration  avg=88ms  p(95)=171mshttp_req_failed  0.00%  ✓ 0  ✗ 14820vus, max  50    iterations=14820passed: threshold http_req_duration p(95)<200passed: threshold http_req_failed rate<0.01
Illustrative output — representative numbers, not a published benchmark.
Live during the runstreaming

THRESHOLD p(95) < 200ms171ms

Virtual users
50 / 50
Latency p(95) vs threshold
171 / 200ms
Error rate vs budget
0.00 / 1.00%

Both thresholds held — the run passes and the release gate stays green.

Setup steps live in the guidebook.

The guidebook carries the current install and configuration steps for every kit, kept in one place so they stay accurate. Open to Techvify users and invited guests.

Open the guidebook

Start

Gate your next release.

Bring one application and its next release. On a scoping call we agree what each dimension checks, what fails the run and who signs, before any work starts.