- Home
- Skills
- Code Quality & Review
- editor to export regression guard
editor to export regression guard
Prevent data loss and semantic regressions when changing visual editors, serializers, or document renderers.
Free
Works with the AI Agents you already use
editor to export regression guard
Example session with this skill installed
I've updated the Markdown serializer to handle nested lists differently. Run the regression guard against the current main branch as a baseline using the files in /docs/examples.
- Read your context and instructions
- Compiled the editor to export
I have completed the regression check between the main branch (baseline) and the local serializer-update branch (candidate). I focused on nested list serialization in /docs/examples/nested-lists.md.
Regression Guard Report
| Section | Detail |
|---|---|
| Target | Baseline: main, Candidate: serializer-update. Change: Updated nested list indentation logic. |
| Behavior Map | UI Entry (Editor) -> AST Normalization -> Disk Save -> HTML/PDF Export. |
| Verdict | REGRESSION |
Findings
- Trigger: Indentation of three-level nested unordered lists.
Expected Behavior: Sub-items should maintain a 2-space offset in the raw .md file to preserve hierarchy on reload.
- Observed Behavior: Candidate flattens third-level items to the second level during the
Savephase. - Location:
src/serializers/markdown.ts(observed via/docs/examples/output_candidate.mdline 42). - Impact: Permanent data loss of document structure upon saving.
Unverified
- PDF export environment was unavailable; verification was performed on raw Markdown and HTML preview artifacts only.
Next steps
- Review the AST normalization logic in
src/serializers/markdown.tsspecifically for depth > 2. - Fix the indentation logic and rerun this guard to move the status from
REGRESSIONtoNO_REGRESSION_OBSERVED. - Provide the PDF generation environment to verify the final output path.
Connects securely to your tools. The creator never sees your data.
About this skill
Changing a document renderer, serializer, or visual editor often introduces silent data loss that unit tests miss. This skill prevents regressions by tracing data from the user's input through storage, reloads, previews, and final exports. It focuses on semantic meaning and content integrity rather than just checking if pixels look similar.
What it does
- Map data journeys by tracing how values move from UI controls to database storage and final output formats.
- Detect semantic regressions where content is lost or meaning changes across different document versions.
- Verify export integrity by inspecting the actual generated files and downloads instead of just testing UI click handlers.
- Normalize comparisons to ignore expected variations like timestamps or generated IDs while flagging critical content shifts.
- Negative control testing to ensure new regression checks actually fail when presented with known defects.
How it works
- Define the target by recording the baseline version, the candidate changes, and the intended differences in document behavior.
- Execute discriminating tests using synthetic or authorized documents to exercise the specific paths most likely to lose data.
- Analyze the behavior map to inspect input normalization, storage formats, and every affected output path.
- Report findings using a strict contract that classifies the result as
REGRESSION,PARTIAL, orNO_REGRESSION_OBSERVED.
Frameworks & tools
This tool is language-agnostic and works with any stack involving document builders, serializers, or renderers. It integrates with your project's local file system and existing check suites to verify JSON, XML, HTML, or binary exports.
Why this beats prompting it yourself
Generic prompts often overlook the "storage-to-reload" cycle where data corruption actually happens. This skill enforces a rigorous behavior map that catches edge cases like empty vs. null states and inconsistent rendering across different export clients.
Use cases
- Upgrading serializers when moving from one document format to another while preserving existing user data.
- Refactoring editor components to ensure new UI controls correctly map to the underlying data schema.
- Validating export engines to confirm that PDF or Markdown generators maintain parity with the live web preview.
Known limitations
Does not perform full accessibility audits or visual screenshot cloning. Requires access to local files or supplied artifacts to establish a baseline.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Code Quality & Review.
Requires accessible local files or artifacts to establish a baseline. Does not perform accessibility audits or visual screenshot matching. Scope limited to provided document paths.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 13 days ago
- Free with an Agensi account