docs(skills): split argent-create-flow into references and add argent-qa-flows

packages/argent-installer/test/skills-frontmatter.test.ts

@@ -0,0 +1,29 @@
+import fs from "node:fs";
+import path from "node:path";
+import { fileURLToPath } from "node:url";
+
+import { parse as parseYaml } from "yaml";
+import { describe, expect, it } from "vitest";
+
+const skillsDir = fileURLToPath(new URL("../../skills/skills/", import.meta.url));
+
+describe("bundled skill frontmatter", () => {
+  it("parses every SKILL.md YAML block", () => {
+    const skillFiles = fs
+      .readdirSync(skillsDir, { withFileTypes: true })
+      .filter((entry) => entry.isDirectory())
+      .map((entry) => path.join(skillsDir, entry.name, "SKILL.md"))
+      .filter((filePath) => fs.existsSync(filePath));
+
+    expect(skillFiles.length).toBeGreaterThan(0);
+    for (const filePath of skillFiles) {
+      const content = fs.readFileSync(filePath, "utf8");
+      const frontmatter = content.match(/^---\r?\n([\s\S]*?)\r?\n---/)?.[1];
+      expect(frontmatter, `${filePath} is missing YAML frontmatter`).toBeDefined();
+      expect(
+        () => parseYaml(frontmatter!),
+        `${filePath} has invalid YAML frontmatter`
+      ).not.toThrow();
+    }
+  });
+});

packages/skills/rules/argent.md

@@ -76,6 +76,7 @@ Decision order:
 - Interaction tools (`gesture-tap`, `gesture-swipe`, `gesture-pinch`, `gesture-rotate`, `gesture-custom`, `launch-app`, etc.) return a screenshot automatically.
   Call `screenshot` separately only for a baseline before any action or after a delay.
 - Always open apps with `launch-app` or `open-url` — never tap home screen icons.
+- If there is any chance the task ends in a recorded flow or QA test, do not interact with the app first: load `argent-create-flow` and start the recorder before the first launch or in-app action. A path already walked cannot be recorded retroactively.
 - Always use `run-sequence` when performing multiple sequential device actions where you don't need to observe the screen between steps. More in `argent-device-interact` skill.
 - When the session ends or the user says they are done: call `stop-all-simulator-servers` with `devices: [...]`
   naming the devices this session actually used. One tool-server is shared by every other agent using this
@@ -131,7 +132,7 @@ When: Explicit visual regression, screenshot diff, compare screenshots, before/a
 
 SCREEN RECORDING (VIDEO CAPTURE)
 Skill: `argent-screen-recording`
-When: The user wants a video of the device screen — recording a flow, interaction, animation, or bug reproduction as a clip, or documenting app behavior beyond what a still screenshot shows. Covers the start → interact → stop lifecycle, the reminder discipline that keeps a recording from being left running, and retrieving the mp4 artifact.
+When: The user wants a video of the device screen — recording a flow, interaction, animation, or bug reproduction as a clip, or documenting app behavior beyond what a still screenshot shows. Covers the start → interact → stop lifecycle, the reminder discipline that keeps a recording from being left running, and retrieving the mp4 artifact. Not for a replayable saved sequence: route that meaning of "record a flow" to `argent-create-flow`.
 Prompt keywords: record, recording, screen recording, video, capture video, clip, mp4
 
 RUNNING / BUILDING / DEBUGGING REACT NATIVE APP
@@ -154,15 +155,21 @@ PERFORMANCE OPTIMIZATION
 Use skill: `argent-react-native-optimization`
 When: App feels slow, user asks to optimize, reducing bundle size, improving startup time, fixing re-renders, optimizing lists/images/navigation, or any performance-related task. This is the entry-point skill for all performance work — it delegates to `argent-react-native-profiler` for measurement.
 
-END-TO-END UI TESTING
+INTERACTIVE UI TESTING (ONE-OFF, NOT SAVED)
 Skill: `argent-test-ui-flow`
-When: Verifying complete user flows, running interact → screenshot → verify loops, testing features by using the app, executing manual QA steps, or validating visible UI changes or visual behavior after implementation.
+When: Verifying complete user flows, running interact → screenshot → verify loops, testing features by using the app, executing manual QA steps, or validating visible UI changes or visual behavior after implementation. Not for a test that must be saved and re-run — see GENERATED QA REGRESSION TESTS below.
 
 RECORDING & REPLAYING FLOWS
 Use skill: `argent-create-flow`
-When: A multi-step interaction sequence needs to be repeated — re-profiling after a fix, A/B comparisons, regression checks, user says "again" / "run that flow", or you worked through a complex path worth saving. Also use proactively: if you are about to repeat steps you already performed, record first, then replay.
+When: A multi-step interaction sequence needs to be repeated — re-profiling after a fix, A/B comparisons, user says "again" / "run that flow", or you worked through a complex path worth saving. Also use proactively: if you are about to repeat steps you already performed, record first, then replay. For a QA test case, ticket, or acceptance criteria to keep as a regression test, use `argent-qa-flows` (it loads this skill as its engine).
 Prompt keywords: flow, repeat, test X times
 
+GENERATED QA REGRESSION TESTS
+Use skill: `argent-qa-flows`
+When: The user gives a test case, ticket, or acceptance criteria to keep as a repeatable test — "generate a QA test", "turn this test case into a flow", "automate this regression check", "make a test that does X and checks Y". Orchestrates `argent-create-flow`, records the first walkthrough live, verifies each requested screen/state with stable evidence, and completes only after the unchanged full flow passes twice consecutively. iOS, Android, and Chromium.
+Prompt keywords: QA test, regression test, test case, automate this test, automate an e2e test, keep this e2e test, generate a test
+Saved-artifact rule: one-off interactive check → `argent-test-ui-flow`; saved replayable path → `argent-create-flow`; saved test with acceptance criteria and two-pass proof → `argent-qa-flows`.
+
 PROPOSING DESIGN VARIANTS FOR HUMAN SELECTION
 Use skill: `argent-lens`
 When: The user asks for design alternatives / options / A-B choices for a screen or component, or you have produced more than one candidate look for an element and want a human to pick before committing. Covers the build → navigate → screenshot → propose_variant loop and the single blocking await_user_selection call. (Gated behind the `argent-lens` flag, off by default — run `argent enable argent-lens` first.)