agentleFS
AhmiDarrowAhmiDarrow/RemedyAI/skills/soak-product/SKILL.md

soak-product

Skill5 starsChanged 24 days agoApache-2.0View on GitHub

Full product soak for Remedy — API health, computer-use, plan mode, browser rail, concurrency, docs check, and sign-off artifacts. Use when soak, product soak, e2e live, computer-use checklist, or pre-merge confidence on desktop+API.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    After computer-use / browser / stream / multi-tab changes

    line 21
  2. Distinctive:

    Before merge of a large feature branch

    line 23
  3. Very specific:

    Update or create docs/full-product-soak-signoff.md (or dated note):

    line 86

≈ “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

Product soak

Goal

End-to-end confidence that Desktop + local API still work as a product — not only unit tests. Produces machine-readable results under docs/.

When to use

  • After computer-use / browser / stream / multi-tab changes
  • User says soak, full product soak, e2e live, sign-off
  • Before merge of a large feature branch

Preconditions

  • API up (http://127.0.0.1:7400 or isolated :7410)
  • Desktop running with computer host when testing CUA
    • Release install or npm run tauri:dev:isolated
  • Build mode (not Plan-only) for click/type
  • Token: ~/.remedy/auth/local_api_token (or $REMEDY_HOME/auth/…)
$env:REMEDY_API = "http://127.0.0.1:7400"
$env:REMEDY_HOME = "$env:USERPROFILE\.remedy"

Automated soak (primary)

From repo root with API + desktop live:

.\.venv\Scripts\python.exe scripts\_full_product_soak.py
  • Writes docs/_full_product_soak_results.json
  • Exit 0 only if no FAIL (SKIP allowed)
  • Covers health, computer-use, plan allowlist, providers, concurrency, docs

Also useful:

.\.venv\Scripts\python.exe scripts\_soak_run.py
.\.venv\Scripts\python.exe scripts\live_full_product_e2e.py
.\.venv\Scripts\python.exe scripts\computer_rail_e2e.py

Computer-use checklist (manual / agent)

Open F1 or:

help_read(id="computer-use-soak")

Canonical chapter: docs/manual/computer-use-soak.md.

Minimum path:

  1. computer_monitors → ≥1 display
  2. computer_screenshot → PNG under home computer/shots/
  3. computer_snapshot → window refs
  4. Browser: navigate in-rail, snapshot eN, click, stop cancels jobs
  5. Offline: navigate does not surprise-open OS browser

Unit neighbors (always before claiming soak)

uv run pytest -q tests/test_computer_use.py tests/test_react_stream.py tests/test_react_policy.py
uv run python scripts/check_docs.py
cd desktop && npm test

Sign-off format

Update or create docs/full-product-soak-signoff.md (or dated note):

MetricValue
PASS / FAIL / SKIPcounts
API version / port
Desktop host connectedyes/no
VerdictPASS or blocked + next fix

Anti-patterns

  • Soak against dead API
  • “PASS” with FAIL in _full_product_soak_results.json
  • Skipping computer host when the change was CUA/browser
  • Running soak scripts against the wrong home/port during dual-instance dogfood

Related skills

  • self-inject — test-gated auto-improve (draft → gate → apply/rollback)
  • gauntlet-security — security-focused gate
  • stress-suite — higher load after soak is green
  • self-dev-loop — sequence
  • project-etiquette — ship after green soak when releasing

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.