<< All versions

Skill v1.0.0

currentAutomated scan96/100
atagon-gmbh/kogiqa-mcp/exploratory-testing
──Details
PublishedOctober 21, 2026 at 09:42 AM
Content Hashsha256:84e03b915dda9016...
Git SHAa49651e5a62e
──Files
Files (1 file, 1.3 KB)
SKILL.md1.3 KBactive
SKILL.md · 29 lines · 1.3 KB

version: "1.0.0" name: exploratory-testing description: Run time-boxed exploratory testing against a web app to find bugs and inconsistencies. Use when asked to poke at an app, stress its forms, or hunt for edge cases.


Exploratory testing

Systematically explore an app for defects within a time box.

Workflow

  1. browser_navigate to the app and map its reachable pages with browser_list_clickable.
  2. For every form you find (browser_list_page_inputs), try boundary and invalid inputs with

browser_type: empty, very long, unicode, emoji, negative numbers, malformed emails, HTML and script fragments.

  1. Exercise interactive controls: dropdowns (browser_select), uploads (browser_upload_file),

drag targets (browser_drag_from_to), and keyboard shortcuts (browser_keypress).

  1. After each action, check for a change in state with browser_has_text, and watch

browser_get_console_log for errors.

  1. Use browser_fuzzing for time-boxed randomized input only after the manual pass, and

browser_check_accessibility for an accessibility sweep.

Reporting

Produce a numbered list of findings. For each: the page, the exact steps, the observed vs. expected behavior, and a severity (blocker / major / minor). Separate "looks broken" from "looks inconsistent but intentional" and flag the latter as a question.

All versions