Dev to main, 16th July, 2026#768
Merged
Merged
Conversation
- Replaces plain GitHub icon with "Star on [icon] [⭐ count]" pill - Star count fetched from GitHub API with 10-min localStorage cache - Formatted as compact (e.g. 1.2k, 12k) — falls back to ★ if API fails - Matches the landing page header treatment for consistency Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…ls guide Restructures the guide into the two ways to run a guardrail: through the gateway (existing content, plus the Monitor and test tabs) and through the Protect SDK (new). Also swaps the rule-based-checks shot for a fresh capture of the current AI-Powered Checks catalog and softens the Overview claim to match what the dashboard actually shows.
Walkthrough of the Evaluations page: the table columns (including the hidden Versions column), the tag categories and how to filter by them, the Filter and Columns controls, and Create evals. Includes an annotated screenshot and three GIFs for tags, filtering, and columns. Removes the Eval logs sidebar item. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…l screenshots Drops the Steps-wizard layout and generic placeholder images for the same H2-per-action pattern the other eval guides use. Real nav path (Settings -> AI Providers -> Custom model tab -> Create custom model), real drawer copy, and live screenshots of both the From-Provider and Configure-Custom-Model forms.
Walkthrough of the Test Evaluation panel on an eval's page: the four ways to feed it data (Custom, Dataset, Tracing, Simulation), how the eval's variables map to your fields (auto-mapped for Custom, mapped per field for real data), and reading the pass/fail result with its explanation. Includes two annotated screenshots. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- Reorder the four data sources to match the UI tab order (Dataset, Tracing, Simulation, Custom) - Add a colon after each bold term and tighten the Custom bullet to one clause - Link the first mention of Dataset, Tracing, and Simulation - Trim the Custom caption so it no longer repeats the mapping paragraph Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace the two Test Evaluation screenshots with uniform captures (both now the same dimensions), keeping the horizontal callouts on the mode tabs and the Pass result. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Walkthrough of an eval's Usage tab: picking a time range to scope everything below it, reading the summary numbers (Runs, Success, Errors, Task Completion Rate) and the volume vs completion-rate chart, and digging into any single run from the evaluation logs, including searching by trace ID. Includes one annotated screenshot. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- List every time range in the selector (adds 6 hrs, Today, Yesterday) instead of implying a contiguous 30 mins-to-7D range - Correct the Source values to Playground, Dataset, or Tracer, or the project name (it is not a simulation) - Correct the Search description: a free-text match across a row's id, input, result, and reason, not a trace-ID lookup; relabel the screenshot callout Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace the Usage tab screenshot with a capture matching the other page images' dimensions, keeping the time-range and free-text search callouts. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
… overview Output Type also returns a category label (Choices), not just pass/fail or a score. Note the tag buckets are the main chips; evals also carry finer tags like Quality, Bias, or Data Privacy. Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
…back Retitle to The Evaluations page to resolve the doc-vs-UI ambiguity. Split the tags double-exposure: the Tags section now explains only what tags mean, and all filtering (tag chips + the tag-filter demo) moves under Find an eval. Add a concrete Filter example that a chip can't express, and sharpen the Built-in evals card to what that page adds. Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
…overview docs(evaluation): Explore playground overview
…test docs(evaluation): Test an eval
The page jumped straight to the Test Evaluation panel. Add an Open the eval you want to test step, with the Evaluations-list screenshot, and trim the now-redundant open-from-playground line in the hero. Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
…usage docs(evaluation): Usage & analytics
Walk the feedback flow on a dataset and in the eval playground, then review corrections on the Feedback tab. Adds annotated screenshots and wires the page into the Evaluation guides nav.
… screenshot - Reword the not-available Note so raw-number metrics don't read as a contradiction of Score evals (which are feedback-eligible) - Relabel the re-scoring screenshot's callout to 'Stores your correction, not a rule change', matching the page's criteria-unchanged point
docs(evaluation): add Collect feedback guide
Drop the banned About heading for a hero lead, cut the unverified save-is-a-connectivity-check claim (create only validates and stores), cut the When to use and Field reference repetition, rename the footer to Keep exploring with four deliberate cards, and match provider names to the UI. Mask every config block in the four screenshots with the house masked-bar convention, keeping the edit, delete, and copy icons visible. Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
- Replace the gateway UI walkthrough and its 10 screenshots with a short text pointer to the Command Center gateway docs - Make the Protect SDK the core of the page: first check, per-language rule tables, verdict handling, input+output guarding, and Protect Flash, all verified against the SDK source - Flag that the Python and TypeScript SDKs accept different metric names
docs(evaluation): rewrite custom models guide to house style
docs(evaluation): Set up guardrails (how-to)
Adding the remaining guides
Add live GitHub star counter to docs header
atharva-bhange
approved these changes
Jul 16, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Pull Request
Description
Describe the changes in this pull request:
Checklist
Related Issues
Closes #<issue_number>