testeiya 0.9.0 → 0.10.1

package/pipelines/pr-review.yml

@@ -1,5 +1,6 @@
 name: pr-review
 description: Review a pull request, propose test cases for it, and sync them to Testomat.io.
+target: pr
 
 gates:
   code-changed:
@@ -7,37 +8,49 @@ gates:
       min_files: 3
       ignore: ["**/*.md", "**/*.lock", "docs/**"]
     message: Fewer than 3 code files changed.
+  testcases-needed:
+    env: [CREATE_TESTCASES]
+    message: 'No test cases created. Enable test case creation via CREATE_TESTCASES env var'
   testomatio:
-    env: [TESTOMATIO, TESTOMATIO_PROJECT_ID]
-    message: No Testomat.io project. Set TESTOMATIO and TESTOMATIO_PROJECT_ID.
+    env: [TESTOMATIO]
+    message: No Testomat.io project. Set TESTOMATIO.
 
 when: [code-changed]
 
+env:
+  # feature/add-login → feature-add-login
+  TESTOMATIO_BRANCH: ${GIT_BRANCH | slugify}
+
+
 stages:
   - id: requirements
+    title: Requirements
+    emoji: 📋
     session: reviewer
-    task: |
-      Work out what this pull request is meant to do. /qa-pr-requirements-analyzer
-      Write the acceptance criteria to criteria.md.
-    outputs: [criteria.md]
+    task: Work out what this pull request is meant to do. /qa-pr-requirements-analyzer
 
   - id: review
+    title: Review
+    emoji: 🔍
     session: reviewer
-    task: |
-      Review this pull request as a QA engineer against criteria.md. /qa-review-pr
-    inputs: [criteria.md]
+    task: Review this pull request as a QA engineer against requirements.md. /qa-review-pr
+    inputs: [requirements.md]
 
   - id: test-cases
+    title: Test cases
+    emoji: 📝
     session: writer
+    when: [testcases-needed]
     task: |
-      Write test cases for this pull request from criteria.md and review.md. /qa-write-test-cases
+      If there is linked issue and it was already reviewed by Testeiya and have linked tests - skip this step
+      Write test cases for this pull request from requirements.md and review.md. /qa-write-test-cases
       Put them in test-cases/, one markdown file per suite.
-    inputs: [criteria.md, review.md]
+    inputs: [requirements.md, review.md]
     outputs: [test-cases/]
 
   - id: sync
-    when: [testomatio]
-    each_file: test-cases/*.md
+    title: Sync to Testomat.io
+    emoji: 🔄
+    when: [testomatio, testcases-needed]
     task: |
       Sync the test cases in ${ITEM} to Testomat.io. /sync-test-cases-with-tms
-      Skip a case that already exists there.

package/pipelines/regression.yml

@@ -1,5 +1,6 @@
 name: regression
 description: Run the tests a change can break, from the coverage map, then explain the failures.
+target: pr
 
 gates:
   coverage:
@@ -22,6 +23,8 @@ env:
 
 stages:
   - id: tests
+    title: Tests
+    emoji: 🧪
     when: [testomatio]
     run: |
       npx -y @testomatio/reporter run --kind detect --warn \
@@ -30,10 +33,11 @@ stages:
     timeout: 30m
 
   - id: failures
+    title: Failures
+    emoji: 🐞
     when: [local, test-command]
     session: triage
     task: |
-      Read the test run in tests.log. For each failure, say if this change caused it,
-      if it is flaky, or if the test is out of date. /debug-fix-failed-flaky-autotests
-      Do not edit any file. Fail the run if this change caused a failure.
+      Triage the failures in tests.log without editing files. /debug-fix-failed-flaky-autotests
+      Fail the run if this change caused a failure.
     inputs: [tests.log]

package/pipelines/web-testing.yml

@@ -1,5 +1,6 @@
 name: web-testing
 description: Review a pull request, test the change in a browser with Explorbot, explain what broke.
+target: pr
 
 gates:
   staging:
@@ -13,38 +14,39 @@ gates:
     message: Explorbot has no AI provider. Set EXPLORBOT_AI_PROVIDER and its key.
 
 env:
-  EXPLORBOT_OUTPUT: ${PIPELINE_DIR}/explorbot
   EXPLORBOT_NO_BANNER: "1"
 
 stages:
   - id: review
+    title: Review
+    emoji: 🔍
     session: qa
-    task: |
-      Review this pull request as a QA engineer. /qa-review-pr
-      Write the user-facing scenarios worth checking in a browser to scenarios.md.
-    outputs: [scenarios.md]
+    task: Review this pull request as a QA engineer. List the scenarios worth checking in a browser. /qa-review-pr
 
   - id: write-plan
+    title: Test plan
+    emoji: 🗺️
     session: qa
     when: [staging]
-    task: |
-      Turn scenarios.md into an Explorbot test plan for ${EXPLORBOT_URL}. /explorbot-plan
-      Write the plan to plan.md. Keep only scenarios a browser can check.
-    inputs: [scenarios.md]
+    task: Turn the browser scenarios in review.md into an Explorbot test plan for ${EXPLORBOT_URL} in plan.md. /explorbot-plan
+    inputs: [review.md]
     outputs: [plan.md]
 
   - id: explore
+    title: Browser testing
+    emoji: 🌐
     when: [staging, browser, explorbot-ai]
-    run: npx -y explorbot test "$PIPELINE_DIR/plan.md" '*'
+    run: EXPLORBOT_OUTPUT="$STAGE_DIR/explorbot" npx -y explorbot test "$STAGE_DIR/plan.md" '*'
     inputs: [plan.md]
     outputs: [explorbot/]
     timeout: 30m
 
   - id: results
+    title: Results
+    emoji: 📊
     session: qa
     when: [staging, browser, explorbot-ai]
     task: |
-      Explorbot results are under explorbot/. Find its session report and read it.
-      Say which failed scenarios are real defects in this change and which are test noise.
+      Read the Explorbot report in explorbot/ and say which failed scenarios are real defects and which are noise. /explorbot-fundamentals
       Fail the run if any is a real defect.
     inputs: [plan.md, explorbot/]

package/prompt/pipeline-report.ts

@@ -0,0 +1,23 @@
+import dedent from "dedent";
+
+export const pipelineReport = dedent`
+  <pipeline-report>
+  The draft is the report: every stage's report, word for word, under its own \`### <emoji> <title>\` heading. Your job is to trim it, never to write it. This format overrides <final-report>, its title included.
+
+  1. Copy the draft to the report file with the \`write\` tool, word for word.
+  2. Trim it with the \`edit\` tool:
+     * Drop a line that repeats what an earlier section already said.
+     * When a section is too long, shorten its lines or drop its least important bullets.
+  3. Nothing repeats and nothing is too long: leave the copy as it is.
+
+  Never:
+  * Add, rename, merge, reorder or drop a heading.
+  * Add a sentence, a title, an intro, a conclusion or a closing summary.
+  * Change a fact, a number, an id or a link. Reword only to shorten.
+  * Cut a numbered list, or replace a list with a count like "10 items, see the file".
+  * Add findings, assumptions or advice of your own.
+
+  * A command stage's section links its log. You may add up to three bullets quoting its failures from the log.
+  * The verdict is given. Do not call a result tool to set one.
+  </pipeline-report>
+`;

package/prompt/report.ts

@@ -18,7 +18,7 @@ export function finalReport(path: string): string {
     <final-report>
     * Write your complete final report to \`${path}\` with the \`write\` tool. Writing it is required before you finish.
     * That file is your answer. It is the run's deliverable; nothing else you say is kept.
-    * A skill's output format overrides every rule in this block, including the title. Follow it strictly.
+    * A skill's or the task's output format overrides every rule in this block, including the title. Follow it strictly.
     * Decide the **severity of your report**. Low severity: a few short bullets, nothing more.
     * Do not make the reader go through a long report for a minor issue with no effect.
     * Severity decides how many items you report, not how long each one is. Even a critical finding is: what breaks, for whom, when.

package/prompt/system-prompt.ts

@@ -15,7 +15,7 @@ export function buildSystemPrompt(options: SystemPromptOptions): string {
   const date = new Date().toISOString().split("T")[0];
   const clis = options.connectedClis?.join(", ") || "none connected yet";
   const mcps = options.connectedMcps?.join(", ") || "none connected yet";
-  let manualTestsHint = " Pulled manual tests usually live in `.testeiya/manual-tests/`; check there before declaring the project has no manual tests.";
+  let manualTestsHint = " Pulled manual tests usually live in `.testeiya/test-cases/`; check there before declaring the project has no manual tests.";
   if (options.manualTestsHint === false) manualTestsHint = "";
 
   const prompt = dedent`

package/prompt/testomat.io.ts

@@ -38,8 +38,8 @@ export function testomatio(options: TestomatioOptions): string {
 }
 
 function dynamicData(tms: TmsAccess): string {
-  if (tms === "mcp-proxy") {
-    return "Get them through the Testomat.io tools, one per entity (`tests`, `runs`, `testruns`, `issues`, …); pick the operation with its `command` param. The `mcp` tool searches them when the name is unclear.";
+  if (tms === "mcp-tools") {
+    return "Get them through the `mcp__testomatio__*` tools, one per entity (`tests`, `runs`, `testruns`, `issues`, …); pick the operation with its `command` param.";
   }
   if (tms === "cli-only") {
     return "There are no Testomat.io tools in this session. Get them from the REST API: `curl` with the `TESTOMATIO` token as the Authorization header against `$TESTOMATIO_URL/api/v2`. Keep to documented endpoints; never invent paths or parameters.";
@@ -48,11 +48,11 @@ function dynamicData(tms: TmsAccess): string {
 }
 
 /**
- * `mcp-proxy`: one tool per entity plus the `mcp` search tool. `cli-only`: no
+ * `mcp-tools`: one tool per entity, named `mcp__testomatio__<tool>`. `cli-only`: no
  * tools at all; `check-tests` and REST through the shell. `mcp-direct`: one tool
  * per entity, prefixed `testomatio-<slug>`; the desktop app's one-shot mode.
  */
-export type TmsAccess = "mcp-direct" | "mcp-proxy" | "cli-only";
+export type TmsAccess = "mcp-direct" | "mcp-tools" | "cli-only";
 
 export interface TestomatioOptions {
   /** How the agent reaches Testomat.io in this run. */

package/skills/skills.lock.json

@@ -79,7 +79,7 @@
     {
       "source": "microsoft/playwright-cli/tree/main/skills/playwright-cli",
       "ref": null,
-      "sha": "b1a5903c5b633b8c9f843ad7b629c2a78437040c",
+      "sha": "86f5616716033854f0990734fe2f9fcde0c1f6cc",
       "folder": "playwright",
       "skills": [
         "playwright-cli"

package/dist/src/mcp-extension.js

@@ -1,9 +0,0 @@
-import { mcpConfig, seedMetadataCache, vendorBundle } from "./mcp.js";
-/** The MCP adapter, configured from the environment, as a pi extension. */
-export default async function mcpExtension(pi) {
-    const config = mcpConfig();
-    await seedMetadataCache(config);
-    const { createMcpAdapter } = (await import(vendorBundle()));
-    await createMcpAdapter({ config })(pi);
-}
-//# sourceMappingURL=mcp-extension.js.map

package/dist/src/mcp-extension.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"mcp-extension.js","sourceRoot":"","sources":["../../src/mcp-extension.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,SAAS,EAAE,iBAAiB,EAAE,YAAY,EAAE,MAAM,UAAU,CAAC;AAEtE,2EAA2E;AAC3E,MAAM,CAAC,OAAO,CAAC,KAAK,UAAU,YAAY,CAAC,EAAW;IACpD,MAAM,MAAM,GAAG,SAAS,EAAE,CAAC;IAC3B,MAAM,iBAAiB,CAAC,MAAM,CAAC,CAAC;IAEhC,MAAM,EAAE,gBAAgB,EAAE,GAAG,CAAC,MAAM,MAAM,CAAC,YAAY,EAAE,CAAC,CAEzD,CAAC;IACF,MAAM,gBAAgB,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,EAAE,CAAC,CAAC;AACzC,CAAC"}