agentleFS
Sign inStart free
AhmiDarrowAhmiDarrow/RemedyAI/skills/stress-suite/SKILL.md

stress-suite

Skill5 starsChanged 24 days agoApache-2.0View on GitHub

Live stress and break suites for Remedy API/desktop — multi-turn storms, concurrent sessions, multi-provider rotation, adversarial break battery. Use when stress, load, break suite, hammer, continuous test, or find weak spots.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    After concurrency / multi-provider / stream abort work

    line 25
  2. Distinctive:

    Prefer release API for partner-grade stress.

    line 32
  3. Distinctive:

    Call out repro steps for any FAIL (session id, provider, last request).

    line 102

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Stress suite

Goal

Push the running local API (and optionally desktop) hard enough to surface races, stalls, auth flakes, and provider-bind bugs that unit tests miss.

Not a substitute for soak or gauntlet — run those first when security/product paths changed.

When to use

  • User says stress, hammer, break suite, continuous test, weak spots
  • After concurrency / multi-provider / stream abort work
  • Nightly or pre-release confidence (with soak green)

Preflight

  1. API healthy: GET {REMEDY_API}/api/ping
  2. Bearer token present under $REMEDY_HOME/auth/local_api_token
  3. Prefer release API for partner-grade stress; use isolated only when testing that build
$env:REMEDY_API = "http://127.0.0.1:7400"
$env:REMEDY_HOME = "$env:USERPROFILE\.remedy"
# optional knobs:
# $env:REMEDY_STRESS_PASSES = "20"

Suites (pick by intent)

A. Full stress battery (API)

.\.venv\Scripts\python.exe scripts\stress_full_suite.py
.\.venv\Scripts\python.exe scripts\live_stress_remedy.py
.\.venv\Scripts\python.exe scripts\stress_desktop_api.py

stress_full_suite.py: multi-pass with provider rotation (demo / deepseek / xai / poe). Honor REMEDY_STRESS_PASSES, REMEDY_STRESS_ROTATE_EVERY.

B. Adversarial break suite

.\.venv\Scripts\python.exe scripts\live_agent_break_suite.py

Abuse-shaped turns, concurrency, policy edges. Stop on first hard FAIL if debugging.

C. Multi-turn / workout

.\.venv\Scripts\python.exe scripts\live_10_turn_stress.py
.\.venv\Scripts\python.exe scripts\live_agent_full_workout.py
.\.venv\Scripts\python.exe scripts\live_continuous_test.py

D. Desktop UI / settings matrix

.\.venv\Scripts\python.exe scripts\live_desktop_ui_10runs.py
.\.venv\Scripts\python.exe scripts\live_settings_matrix.py

Needs desktop automation hooks / live UI — skip if headless CI only.

E. Security-adjacent live chat soak

.\.venv\Scripts\python.exe scripts\live_soak_security_chat.py

How to run safely

RuleWhy
Start small (REMEDY_STRESS_PASSES=5) then scaleFind fail fast
One stress driver at a timeAvoid false races from two scripts
Watch stall banner / logsQuiet streams ≠ always hung
Do not force-quit release partner mid-dogfoodUse isolated profile for WIP

Report format

SuitePassesFailsWeakNotes
stress_full_suite
break_suite
…

Call out repro steps for any FAIL (session id, provider, last request).

Anti-patterns

  • Stress before unit + soak when you just changed jails or CUA
  • Blaming providers for bugs that are session-bind races
  • Leaving 50-pass storms running overnight without log capture
  • Hitting non-loopback APIs

Related skills

  • soak-product — functional green before load
  • gauntlet-security — auth/jail first
  • self-inject — test-gated auto-improve
  • self-dev-loop — orchestration

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free