MahmulPMahmulP/qa-kit/skills/qa-memory/SKILL.md

qa-memory

Skill0 starsChanged 18 days agoMITView on GitHub

Read, write, correct, and consolidate QA memory - per-project knowledge files and searchable lessons in .qa-kit/memory, plus global lessons shared by every app in ~/.qa-kit - so every agent and session keeps learning about the apps and their tests. Use when the user says remember/forget something, when a reviewer corrects a test, at the end of any QA task, or when memory looks stale or contradictory.

1 other project uses a version of this.

Adopted by

Most recently taken up by

fishzjp/qa-skillstheir own version, first seen Oct 21, 2026

Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
fishzjp/qa-skills★ 62same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository5 other files it gives its agents.
- qa-automationSkill
- qa-explore-appSkill
- qa-playwright-authorSkill
- qa-run-triageSkill
- qa-test-casesSkill

The file

We split this skill into the rules it gives an agent

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

≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Very specific:
Do not save: what the code or scan already says plainly, one-off task state, secrets (passwords, tokens, cookies, session files), personal …
line 49
Very specific:
Write each entry as one or two self-contained sentences with evidence (file:line, test ID, run date, ticket).
line 52
Distinctive:
Before your final message, ask yourself: what did I learn that is not in memory yet?
line 80

≈ “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

QA memory

What lives where

Store	Holds	How to change
`.qa-kit/config.json`	Detected conventions: test dir, fixtures import, catalog names/ID pattern/languages, scopes, mutation tags, protected env vars	`qa-kit config set <path> <value>`
`.qa-kit/memory/app.md`	Modules, routes, auth/roles, environments, UI patterns, fragile areas	Edit the file
`.qa-kit/memory/conventions.md`	How tests are written here, folder layout (+ review feedback log)	Edit the file
`.qa-kit/memory/glossary.md`	UI terms and meanings	Edit the file
`.qa-kit/memory/<topic>.md`	Any other curated topic you need (e.g. `test-data.md`)	Create it
`.qa-kit/memory/lessons.jsonl`	Atomic, searchable project entries with kind + scope (IDs `M-…`)	`qa-kit memory add/forget` or `qa_memory_add/forget`
`~/.qa-kit/lessons.jsonl`	Global lessons shared by every app on this machine (IDs `G-…`)	`qa-kit memory add --global`
`.qa-kit/index/`	Derived from code (structure, test-case index)	`qa-kit scan` / `qa-kit index`, never by hand
`.qa-kit/history/`	Run results per test (local, ignored by git)	`qa-kit run` / `qa-kit ingest`

.qa-kit/config.json and .qa-kit/memory/ are meant to be committed so the whole team (and every harness) shares them.

Recall

At the start of a task: qa-kit status, then qa-kit memory search "<words>" --scope <path> for the exact area. Scope matching uses path prefixes, so e2e/billing also returns entries saved for pages under it. Search always includes global lessons.

Project or global?

Project (default): anything about this app — pages, data, selectors, bugs, rules.
Global (--global): true for every app you test — Playwright/tooling behaviour, the team's review habits, CI quirks shared across repos. Never put app-specific facts there. qa-kit apps lists the apps on this machine that share these lessons.

What is worth saving

Save it when the next agent would otherwise lose time or make the same mistake:

a reviewer correction or a user decision (convention, decision),
a locator that works for a tricky control, or one that looked right but is not (selector),
a cause of wasted time: environment quirk, slow page, session conflict (gotcha),
a confirmed or suspected bug with evidence (bug), a flaky test and its cause (flaky),
data that must exist / how to create it (data), how the app behaves (fact).

Do not save: what the code or scan already says plainly, one-off task state, secrets (passwords, tokens, cookies, session files), personal data, long code.

Write each entry as one or two self-contained sentences with evidence (file:line, test ID, run date, ticket). Use the narrowest true scope.

qa-kit memory add --kind convention --source "review 2026-10-24" \
  "List pages need a search-miss test that asserts the empty-state text."
qa-kit memory add --kind bug --scope e2e/billing/invoices \
  "INV-06: export reads row.total instead of row.amount.total (invoice-table.tsx:84-91)."
qa-kit memory add --global --kind gotcha \
  "Playwright's html reporter folder must not be inside outputDir, or each run wipes the report."

Saving the same text again does not duplicate it; it raises its seen count, which ranks it higher in recall and in the session digest. Repetition is a signal: things saved three times belong in the curated .md files.

Correct and consolidate

Wrong or obsolete entry: qa-kit memory forget <ID> --reason "…".
Changed fact: add the new one with --supersedes <old ID>.
Wrong detected convention: qa-kit config set … (it is never overwritten by later scans once set).
Periodically (or when status shows many lessons): read qa-kit memory list, merge repeated lessons into the right .md section, retire the merged entries, and remove sections in the .md files that the code now contradicts. Keep files short; link sources.

End of every QA task

Before your final message, ask yourself: what did I learn that is not in memory yet? Save it, then tell the user in one line what you saved (IDs). If nothing new, say so.

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