How do I install the Opp Repl Statistical Tests skill?

Run npx skillmds@latest add tabgab/opp-repl-statistical-tests in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Opp Repl Statistical Tests safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Opp Repl Statistical Tests?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Opp Repl Statistical Tests free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Opp Repl Statistical Tests?

tabgab (@tabgab) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Opp Repl Statistical Tests by tabgab

Name: Opp Repl Statistical Tests
Author: tabgab

Opp Repl Statistical Tests

Detect regressions in simulation SCALAR results by comparing against saved baselines. Complementary to fingerprint tests — coarser (numeric tolerances) but easier to interpret. update_statistical_test_results seeds or refreshes the baseline in the project's statistics_folder. Load when fingerprint churn is too noisy and you only care about headline metrics.

tabgab d079a35 2.6 KB Updated Sep 17, 2026

File contents

Statistical tests

Statistical tests compare current scalar results (.sca values) against stored baselines. They tolerate numerical noise up to a configured threshold, so they are stabler than fingerprint tests for stochastic simulations but weaker at catching subtle bugs.

Upstream reference: https://github.com/omnetpp/opp_repl/blob/main/doc/statistical_tests.md

Baseline location

SimulationProject parameter statistics_folder (default ".") names the folder where baseline scalar results live, one file per config/run.

Python API

# Seed / refresh after intentional changes
update_statistical_test_results(simulation_project=inet_project)

# Run the tests
run_statistical_tests(simulation_project=inet_project)

# Scope to one region of the project
run_statistical_tests(simulation_project=inet_project,
                      working_directory_filter="examples/ethernet")

Result codes follow the UpdateTask family: KEEP / INSERT / UPDATE on updates, PASS / FAIL on tests.

Command line

opp_update_statistical_test_results --load inet.opp -p inet
opp_run_statistical_tests          --load inet.opp -p inet

When to pick statistical over fingerprint

The simulation has non-determinism you can't remove; fingerprint tests flap but scalar averages are stable within tolerance.
You only care about a handful of headline metrics (throughput, delay, PER).
Your CI budget doesn't allow eventlog-level checks.

Conversely, for tight regression coverage (each event matters), prefer fingerprint tests — see opp-repl-fingerprint-tests.

Pitfalls

Baselines live OUTSIDE the git repo by default (the folder is .). Point statistics_folder at tests/statistics/ or similar to keep the baseline under version control.
Tolerances are per-test and default to vendor values; consult the live docstring (help(run_statistical_tests)) when tuning.
Updating a baseline should be a reviewable commit — otherwise regressions can slip in under cover of a "refresh".

---name: opp-repl-statistical-testsdescription: Detect regressions in simulation SCALAR results by comparing against saved baselines. Complementary to fingerprint tests — coarser (numeric tolerances) but easier to interpret. update_statistical_test_results seeds or refreshes the baseline in the project's statistics_folder. Load when fingerprint churn is too noisy and you only care about headline metrics.---# Statistical testsStatistical tests compare current scalar results (`.sca` values)against stored baselines. They tolerate numerical noise up to aconfigured threshold, so they are stabler than fingerprint testsfor stochastic simulations but weaker at catching subtle bugs.Upstream reference:https://github.com/omnetpp/opp_repl/blob/main/doc/statistical_tests.md## Baseline location`SimulationProject` parameter `statistics_folder` (default `"."`)names the folder where baseline scalar results live, one file perconfig/run.## Python API # Seed / refresh after intentional changes update_statistical_test_results(simulation_project=inet_project) # Run the tests run_statistical_tests(simulation_project=inet_project) # Scope to one region of the project run_statistical_tests(simulation_project=inet_project, working_directory_filter="examples/ethernet")Result codes follow the UpdateTask family:`KEEP` / `INSERT` / `UPDATE` on updates, `PASS` / `FAIL` on tests.## Command line opp_update_statistical_test_results --load inet.opp -p inet opp_run_statistical_tests --load inet.opp -p inet## When to pick statistical over fingerprint- The simulation has non-determinism you can't remove; fingerprint tests flap but scalar averages are stable within tolerance.- You only care about a handful of headline metrics (throughput, delay, PER).- Your CI budget doesn't allow eventlog-level checks.Conversely, for tight regression coverage (each event matters),prefer fingerprint tests — see `opp-repl-fingerprint-tests`.## Pitfalls- Baselines live OUTSIDE the git repo by default (the folder is `.`). Point `statistics_folder` at `tests/statistics/` or similar to keep the baseline under version control.- Tolerances are per-test and default to vendor values; consult the live docstring (`help(run_statistical_tests)`) when tuning.- Updating a baseline should be a reviewable commit — otherwise regressions can slip in under cover of a "refresh".## See also- `opp-repl-fingerprint-tests` — trajectory-level alternative.- `opp-repl-running-simulations` — underlying run machinery.- `opp-repl-tasks-and-results` — inspect FAIL details.

Frequently asked questions