discouraging use of screenshots for verification

[angiejones]

Summary by CodeRabbit

• Chores

  • Package version bumped to 0.2.3

• Documentation

  • Clarified guidance for server-side verification tools: prefer attribute checks or scripted batch reads for state validation, and reserve screenshots for visual/layout-only verification to improve efficiency and reliability.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

---

📝 Walkthrough

Walkthrough

The PR bumps package version to 0.2.3 and expands the inline descriptions for three server-side tools—take_screenshot, get_element_attribute, and execute_script—to clarify recommended usage, batching, and when to prefer attribute/script checks over screenshots.

Changes

Cohort / File(s)	Summary
Version Bump package.json	Version increment from 0.2.2 to 0.2.3.
Server tool docs src/lib/server.js	Expanded descriptions for take_screenshot, get_element_attribute, and execute_script to clarify use cases (when to capture screenshots vs. read attributes), suggest batching reads via execute_script, and recommend preferring non-visual checks for state verification.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~2 minutes

Poem

🐰 I hopped through code with gentle cheer,

I nudged the version, made guidance clear,
"Snap when needed, read attributes fast,"
A little rabbit's changelog cast,
0.2.3 — onward we steer.

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'discouraging use of screenshots for verification' accurately reflects the main changes: expanding descriptions of take_screenshot, get_element_attribute, and execute_script to discourage screenshots for verification purposes.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

• 📝 Generate docstrings (stacked PR)
• 📝 Generate docstrings (commit on current branch)

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch accessibility-tip

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

src/lib/server.js (1)

439-439: Consider also referencing get_element_text for text-content verification.

The description steers agents toward get_element_attribute and execute_script, but omits get_element_text, which is the most direct tool for verifying visible text — likely the most common reason an LLM would reach for a screenshot. Adding it alongside get_element_attribute closes that gap.

✏️ Suggested description update

-        description: "captures a screenshot of the current page. Prefer using the accessibility://current resource for understanding page content. Use get_element_attribute or execute_script to verify element state. Only use screenshots when visual layout or styling needs to be verified.",
+        description: "captures a screenshot of the current page. Prefer using the accessibility://current resource for understanding page content. Use get_element_text to verify text content, get_element_attribute to verify element state, or execute_script to batch-read multiple values. Only use screenshots when visual layout or styling needs to be verified.",

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@src/lib/server.js` at line 439, The description string for the screenshot
tool omits recommending get_element_text for verifying visible text; update the
description (the string assigned to description in the screenshot tool
definition) to mention get_element_text alongside get_element_attribute and
execute_script as the preferred way to verify text content before taking
screenshots so agents use get_element_text for text verification rather than
defaulting to screenshots.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@src/lib/server.js`:
- Line 439: The description string for the screenshot tool omits recommending
get_element_text for verifying visible text; update the description (the string
assigned to description in the screenshot tool definition) to mention
get_element_text alongside get_element_attribute and execute_script as the
preferred way to verify text content before taking screenshots so agents use
get_element_text for text verification rather than defaulting to screenshots.