> For the complete documentation index, see [llms.txt](https://learning.contextqa.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://learning.contextqa.com/reference/changelog.md). # Changelog Changelog for ContextQA documentation — tracks additions, updates, platform release notes, and MCP server version history. ## Quick answer Changelog for ContextQA documentation — tracks additions, updates, platform release notes, and MCP server version history. Use this page to identify documentation changes and then follow the linked feature guide for current user instructions. ## What this page covers This page tracks all significant additions, updates, and restructuring of the ContextQA documentation. For platform release notes and MCP server version history, see the sections below. *** ## 2026-09-08 — Expanded interactive demo coverage * Added focused Storylane walkthroughs for API testing, API test creation, API chaining, hybrid API-and-UI automation, mobile testing, mobile gestures, self-healing, conditional validation, conditions, loops, variables, environments, and the overall platform workflow. * Replaced older 69–96-step platform, mobile, variable, and environment embeds with shorter feature-specific walkthroughs where available. * Created dedicated GitBook share links so documentation engagement can be distinguished from other Storylane traffic. * Enabled guided pointers and tracked **Talk to Sales** actions without placing an email gate before the learning experience. * Added crawlable, citation-ready descriptions and outcome summaries around every embed so people, search engines, and AI assistants can understand the feature even when the interactive player is not rendered. * Completed course-section coverage for test suites and plans, first mobile execution, mobile plans, AI root-cause and impact analysis, dynamic data generation, Salesforce automation, knowledge-base context, and Jira defect reporting. * Added split-course navigation for long lessons so Part 1 and Part 2 remain ordered while Storylane reports their engagement separately. * Replaced the remaining default Jira and Salesforce embeds with dedicated GitBook links and removed the obsolete Salesforce email-gate explanation. *** ## 2026-09-07 — Feature-specific interactive demos * Embedded published Storylane walkthroughs on the platform overview, mobile test generation, Jira integration, environment management, and test data management guides. * Added three focused tours—requirements to generated tests, execution evidence, and administration governance—using only three or four current product screens per demo. * Added crawlable, citation-ready descriptions and outcome summaries around every embed so the workflow remains understandable when a search engine or AI agent cannot render the interactive player. * Standardized the demos on guided pointers and a tracked sales action without requiring an email before the walkthrough begins. * Replaced two older shared links with the current environment and variable demos, and expanded the interactive-demo index so visitors can choose a walkthrough by role and desired outcome. *** ## 2026-09-03 — Analytics-led navigation and URL recovery * Reviewed three months of GitBook page, AI-agent, and broken-URL analytics to prioritize changes by observed visitor demand. * Added version-controlled redirects for high-traffic legacy documentation routes covering test creation, execution history, environments, integrations, mobile testing, API testing, self-healing, failure analysis, and file uploads. * Expanded the documentation home-page task index and the canonical AI-agent topic map around the workflows most frequently opened by people and retrieval-enabled agents: mobile builds, execution evidence, AI test generation, API chaining, flaky-test analysis, MCP, and access control. * Added stronger related-page navigation, a contextual demo action, and an explicit correction-feedback path to the mobile upload guide. * Left ambiguous removed topics, malformed URLs, and security-probe paths as real 404s instead of redirecting them to unrelated content. *** ## 2026-09-03 — Content-gap remediation * Reviewed all 31 open GitBook content-gap findings and updated the canonical user-facing pages instead of creating parallel or conflicting guides. * Added safe credential handling for knowledge bases, environment-backed secrets, public interactive demos, authenticated crawls, OTP workflows, and CAPTCHA boundaries. * Corrected reporting guidance for authenticated run links, Execution History CSV exports, AI Insights activity deltas, heatmap availability, root-cause evidence, and retry-based flakiness triage. * Expanded MCP and API documentation with current ChatGPT custom-app prerequisites, plan-execution response fields and mixed outcomes, DAST scope limits, REST contract availability, and API Data JSON-path selection. * Clarified recorder versus runner behavior, stable Salesforce Lightning locators, deterministic DOM-text capture, value-source interpolation, AI Ask-to-Custom Code handoff, calculation assertions, and empty step-group creation. * Aligned Requirements and Jira terminology with the current navigation, documented badge meanings, and added recovery guidance for invalid Jira responses and missing screen or HTML capture data. * Kept this release documentation-only; it does not describe internal business logic or imply unsupported product capabilities. *** ## 2026-09-01 — Production documentation alignment * Updated API-token documentation for exact collection and item routes, the **View System Audit** scope, and exact IPv4 or IPv6 allowlists; added a current production screenshot and focused FAQs. * Corrected requirements documentation to match the production information architecture: six contextual tabs, **Requirement Data** sub-tabs, background generation behavior, and inline links for cases that already exist in the repository. * Added a text-backed Mermaid workflow from requirement source through duplicate review and approval or auto-publish so people, search engines, and retrieval-enabled AI assistants can understand the complete decision path. * Clarified that mobile run results identify the build snapshot used by that execution, live duration survives reconnects and event gaps, and the targeted plan rerun includes failed and not-executed cases. * Documented how user-provided values are applied to AI-suggested fixes before the repaired step is re-run. * Kept this entry documentation-focused; internal implementation details and non-user-facing fixes are intentionally excluded. *** ## 2026-08-28 — Production workflow documentation * Added a citation-ready guide to the limited-rollout **PR Impact Analysis** experience, including repository scope, confidence, read-only evidence, advisory GitHub reporting, superseded and partial analyses, and an end-to-end workflow diagram. * Added organization **Proxy** and **Remote Browser** profile instructions with current production screenshots, secret-handling behavior, remote connection validation, plan and test-case assignment, and a selection-precedence diagram. * Updated requirements generation for the integration chooser, Azure DevOps work-item selection, valid Swagger/OpenAPI JSON inputs, plan-level connection profiles, and AI-ready crawl goals. * Expanded Azure DevOps documentation from defect creation alone to the full work-item-to-requirement and failure-to-defect loops. * Updated test results for the Jira/Azure DevOps bug destination chooser, mobile build names and downloads, and the on-premises-only **Debug log** action. * Removed obsolete test-plan **Run Manually** instructions and standardized plan execution on **Run Now**. * Documented knowledge-base deletion safeguards, missing-reference behavior, bulk test-case status updates, and supported While Loop child-step behavior. * Added current production screenshots, descriptive alternative text, focused FAQs, internal links, concise Quick answers, and contextual demo calls to action for improved search and AI retrieval. *** ## 2026-08-22 — Repository-wide AI readability review * Reviewed all 131 pages published in the documentation navigation. * Added a concise, page-specific **Quick answer** to every published page so search engines and retrieval-enabled assistants can identify the primary answer near the top. * Added related-documentation links to pages that previously had no internal navigation path. * Standardized the public conversion action on **Book a Demo** and removed stale **Start Free Trial** links from documentation calls to action. * Updated the Chrome extension link to its current canonical Chrome Web Store destination. * Expanded the documentation audit to require Quick answers, related-page links, unique titles, unique discovery descriptions, unique answers, and valid heading hierarchy across every published page. * Rechecked all public videos, AI discovery endpoints, MCP guides, calls to action, and third-party instructional links. *** ## 2026-08-21 — User-facing workflow updates * Documented the organization execution-capacity card on Test Plans and Schedules, including running, capacity, queued, and auto-refresh states. * Added the plan-level **Live execution** workflow, **Stop Run**, run-history navigation, and zero-test safeguards. * Added bulk **Change Status** instructions and clarified that the latest-result filter contains settled outcomes only. * Documented mobile app-build names and build downloads in run results. * Added **Created by** attribution to test-case inspection guidance. * Added Excel workbook generation to Document Generation steps. * Clarified requirement-variable sources, value precedence, duplicate handling, syntax review, and environment metadata. * Explained how to clear the organization default knowledge base with **None**. * Added citation-ready **Quick answer** sections and more specific discovery descriptions to the updated workflow pages. * Verified the public crawler, page-sitemap, `llms.txt`, Markdown, canonical URL, and GitBook grounded-query surfaces. * Added reusable retrieval instructions for ChatGPT, Claude, and other URL-enabled assistants, plus automated checks for concise near-top answers on high-intent pages. *** ## 2026-08-12 — Search and AI discoverability * Added a public guide for using the documentation with search engines and AI agents, including the sitemap, `llms.txt`, Markdown pages, canonical citations, and GitBook's grounded-answer query format. * Added direct-answer summaries and query-focused FAQs to high-intent pages for the product, web testing, API testing, AI Insights, self-healing, administration, and onboarding. * Added searchable written companions to the video library so each recording's purpose and workflow remain understandable without video playback. * Added documentation-audit requirements for published-page titles, descriptions, description length, and a level-one heading. * Added contextual demo calls to action on the mobile, AI Insights, video, business-use-case, and MCP overview pages, all using the canonical `https://contextqa.com/book-a-demo/` destination. * Extended the documentation audit to require exactly one level-one heading, descriptive image text, the canonical demo URL, and CTA coverage on high-intent overview pages. * Clarified that GitBook's generated crawler and AI discovery endpoints remain the canonical site-level sources. *** ## 2026-03-15 — Full Documentation Rewrite ### Summary A senior technical writer agent performed a comprehensive rewrite of all ContextQA documentation. The process involved reading every file in the MCP server repository, mapping the complete Angular UI routing tree, executing live test cases for evidence capture, and cross-referencing all features against both code and UI sources. *** ### New Pages Added **Getting Started** * `getting-started/introduction.md` * `getting-started/quickstart.md` * `getting-started/core-concepts.md` * `getting-started/architecture-overview.md` **Web Testing** * `web-testing/creating-test-cases.md` * `web-testing/test-steps-editor.md` * `web-testing/managing-test-suites.md` * `web-testing/test-data-management.md` * `web-testing/self-healing.md` **Execution** * `execution/running-tests.md` * `execution/scheduling.md` * `execution/environments.md` **Reporting** * `reporting/test-results.md` **Integrations** * `integrations/jira.md` * `integrations/github-actions.md` **MCP Server** * `mcp-server/overview.md` * `mcp-server/installation-and-setup.md` * `mcp-server/authentication.md` * `mcp-server/agent-integration-guide.md` * `mcp-server/tool-reference/README.md` **AI Features** * `ai-features/ai-test-generation.md` * `ai-features/autonomous-agent-pipeline.md` **Administration** * `administration/roles-and-permissions.md` * `administration/team-management.md` **Reference** * `reference/glossary.md` * `reference/changelog.md` *** ### Research Completed **MCP Server Analysis** * Read every Python file in the `cqa-mcp` repository * Catalogued all 67 tools with names, descriptions, and parameters * Identified 14 tool categories and their purpose * Documented the authentication flow (per-request login, no session caching) * Documented three deployment options: local uv, Docker, Google Cloud Run * Documented credential resolution priority order **UI Route Map** * Mapped 55 Angular routes across all feature areas * Documented the `LockDataGuard` pattern and which features it gates * Traced workspace version ID usage in URL paths * Identified key navigation patterns (workspace switcher, test development, settings) **Feature Coverage Matrix** * Compared MCP tool coverage against UI feature coverage * Identified features accessible only via UI (certain settings pages, mobile provisioning) * Identified features accessible only via MCP (test repo analysis, migration) * Cross-referenced generation source tools against documented UI flows **Live Execution Evidence** * Executed 10 test cases across different application types * Verified execution polling patterns and timing characteristics * Confirmed evidence package artifacts: screenshots, video, HAR, console, trace * Verified AI root cause analysis output format * Confirmed self-healing suggestion format and confidence threshold behavior *** ### Issues Found See `research/discrepancies.md` for 10 identified discrepancies between the code and documentation, including: 1. The authentication model (per-request re-login vs. cached session) was undocumented — documented for the first time in this rewrite 2. The `get_test_step_results` tool vs. `get_execution_step_details` distinction was unclear — both documented with use case guidance 3. n8n workflow generation tool name (`generate_contextqa_tests_from_n8n`) differed from the display name used in UI — documented the canonical tool name 4. The credential resolution priority order (query params → env vars → .env file) was not documented anywhere — added to authentication page 5. `export_to_playwright` and `export_test_case_as_code` both produce Playwright output but with different scope — clarified: `export_test_case_as_code` is per-test-case, `export_to_playwright` is workspace-wide *** ### Method This documentation was produced by a senior technical writer agent that: 1. Read every file in the MCP server repository (`app/fastmcp_server.py`, `app/contextqa_client.py`, `app/tools/*.py`, `README.md`, `docker-compose.yml`, `Dockerfile`) 2. Read the complete Angular UI routing tree to map all 55 platform routes 3. Executed 10 live ContextQA test cases and captured evidence for documentation 4. Cross-referenced all 67 MCP tools against the ContextQA REST API and UI features 5. Wrote 14 documentation pages covering MCP server, AI features, reporting, integrations, administration, and reference 6. Verified all code examples by tracing them through the implementation code *** ## MCP Server v1.0.0 **Released:** 2025 Initial public release of the ContextQA MCP Server with 67 tools across 14 categories. ### Tools Released **Test Case Management (8 tools)** * `create_test_case` — create test cases from a URL and natural language description * `get_test_cases` — list and filter test cases in a workspace * `get_test_case_steps` — retrieve complete step definitions for a test case * `update_test_case_step` — modify an individual step * `delete_test_case_step` — remove a step from a test case * `delete_test_case` — permanently delete a test case and its history * `query_contextqa` — semantic search across the test case library * `create_complex_test_step` — add conditional, loop, API call, or custom code steps **Execution & Results (5 tools)** * `execute_test_case` — trigger a single test case execution * `get_execution_status` — poll for completion status * `get_test_case_results` — retrieve the full result object with evidence URLs * `get_execution_step_details` — step-by-step breakdown with screenshot URLs * `fix_and_apply` — end-to-end failure detection and fix pipeline **Test Suites & Plans (6 tools)** * `get_test_suites` — list all test suites * `execute_test_suite` — run all test cases in a suite * `get_test_plans` — list all test plans * `execute_test_plan` — trigger a full plan execution * `get_test_plan_execution_status` — poll plan-level execution status * `rerun_test_plan` — re-run a previously executed test plan **Infrastructure & Config (8 tools)** * `get_environments` — list all configured environments * `get_test_devices` — list available mobile device configurations * `get_mobile_concurrency` — check available mobile execution slots * `get_ui_elements` — access the element repository for a page * `list_custom_agents` — list all custom AI agent personas * `create_custom_agent` — define a new agent persona with a system prompt * `list_knowledge_bases` — list all knowledge bases * `create_knowledge_base` — create a new knowledge base with AI instructions **Test Data Profiles (5 tools)** * `get_test_data_profiles` — list all data profiles * `get_test_data_profile` — get the full content of a profile * `create_test_data_profile` — create a new parameterized data profile * `update_test_data_profile` — modify rows and columns * `delete_test_data_profile` — delete a data profile **Test Generation (10 tools)** * `generate_contextqa_tests_from_n8n` — generate from n8n workflow files or URLs * `generate_tests_from_code_change` — generate from a git diff * `generate_tests_from_jira_ticket` — generate from Jira or Azure DevOps tickets * `generate_tests_from_linear_ticket` — generate from Linear issues * `generate_tests_from_figma` — generate from Figma design URLs * `generate_tests_from_requirements` — generate from plain text requirements * `generate_tests_from_excel` — generate from Excel or CSV test libraries * `generate_tests_from_swagger` — generate from OpenAPI specifications * `generate_tests_from_video` — generate from screen recording videos * `generate_edge_cases` — AI-inferred boundary and negative test scenarios **Bug & Defect (3 tools)** * `create_defect_ticket` — push a test failure to Jira or Azure DevOps * `get_auto_healing_suggestions` — AI-proposed locator fixes for failed steps * `approve_auto_healing` — accept and apply a healing suggestion **Advanced Testing (3 tools)** * `execute_performance_test` — run a load or performance test * `execute_security_dast_scan` — run a DAST security scan * `export_test_case_as_code` — export a test case as runnable code (Playwright TypeScript) **AI-Powered Analysis (3 tools)** * `get_root_cause` — AI root cause analysis of a specific test failure * `query_repository` — semantic search of the test repository * `analyze_test_impact` — identify tests impacted by a code change **Analytics & Coverage (2 tools)** * `analyze_coverage_gaps` — identify application flows with no test coverage * `generate_tests_from_analytics_gap` — create tests to close identified gaps **Custom Agents & Knowledge Bases (4 tools)** * `list_custom_agents` — list all custom agent personas * `create_custom_agent` — create a new custom agent * `list_knowledge_bases` — list all knowledge bases * `create_knowledge_base` — create a new knowledge base **Telemetry (5 tools)** * `get_test_step_results` — raw per-step result data * `get_network_logs` — browser HAR network log for an execution * `get_console_logs` — browser console output for an execution * `get_trace_url` — Playwright trace viewer URL * `get_ai_reasoning` — per-step AI confidence scores and locator decisions **Support-to-Fix (2 tools)** * `reproduce_from_ticket` — reproduce a bug from a support ticket * `investigate_failure` — deep investigation of a specific execution failure **Migration Platform (3 tools)** * `analyze_test_repo` — analyze a test repository and report its structure * `migrate_repo_to_contextqa` — convert existing test code to ContextQA * `export_to_playwright` — export all ContextQA tests as Playwright TypeScript ### Configuration Variables | Variable | Required | Description | | -------------------- | -------- | -------------------------------------------------- | | `CONTEXTQA_USERNAME` | Yes | ContextQA account email address | | `CONTEXTQA_PASSWORD` | Yes | ContextQA account password | | `N8N_API_KEY` | No | API key for n8n Cloud workflow integration | | `CONTEXTQA_TENANT` | No | Tenant identifier for multi-tenant n8n deployments | ### Server Configuration * **Default port:** 8080 * **Health check endpoint:** `GET /health` * **MCP endpoint:** `POST /mcp` * **Supported transports:** HTTP (default), stdio, SSE * **Authentication:** Per-request re-login using CONTEXTQA\_USERNAME + CONTEXTQA\_PASSWORD * **Minimum Python version:** 3.9 *** ## Related Pages * [MCP Server Overview](/mcp-server/overview.md) * [Installation & Setup](/mcp-server/installation-and-setup.md) * [Authentication](/mcp-server/authentication.md) * [Tool Reference](/mcp-server/tool-reference.md) --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://learning.contextqa.com/reference/changelog.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.