Evaluation revamp docs#719
Open
khushalsonawat wants to merge 40 commits into
Open
Conversation
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
docs(evaluation): Eval types (three types)
…s/eval-composite-evals
docs(evaluation): Composite evals & groups (concept)
Repoints the judge-model card to concepts/judge-models, where that content now lives after a84dce3 folded it in.
docs(evaluation): Overview revamp
…ation docs(evaluation): Understanding Evaluation concept
…e-agi/docs into docs/eval-custom
What context is (the grounding an eval checks against), the RAG/retrieval evals that read it, knowledge base vs context, agent multi-hop context, and how to provide it. Traced runs still map the span attribute to the eval input (not auto-mapped). No About, minimal frontmatter, no em-dashes, cols=3, all links resolve. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
… guides/ Review fixes on the custom eval page: the Model step lists all three options from the product UI (Use Future AGI Agents / Use other LLMs / Function based), output types and optional settings set as bullets, SDK response comment shows the real envelope, footer cards flow forward. Move the four feature pages to the sidebar's guides/ structure (running-evaluations, custom-evals, custom-models, cicd), sync titles to the sidebar labels, and repoint every reference directly with no new redirect entries. The Overview hero now calls out audio-native scoring. Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
docs(evaluation): Create a custom eval (Khushal-align)
…s/eval-running-evaluations
Open directly at 'What context is' with no lead paragraph, shorten the description, and link the dataset first mention. No diagram (a flow read as forced here). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Sync the title to the sidebar label, repoint stale features/ links to guides/, drop the invented stored-baseline claim on the CI/CD row, add the template-to-surfaces fan-out diagram, define offline/online in place, and link the playground row to the Explore playground overview. Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
docs(evaluation): Running evaluations on Future AGI (hub)
…e-agi/docs into docs/eval-context
Context is an input some templates declare (like output or query), not an object in the evaluation model. Remove the page and its sidebar entry; the grounding idea folds into Eval templates & versions later. Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
docs(evaluation): Context concept
…e-agi/docs into docs/eval-guardrails
A guardrail is a safety-valve eval that blocks content from reaching the user; the two run paths are the Agent Command Center gateway (built-in, third-party, and native Protect checks) and the Protect SDK. Drop the wrong Protect-gateway naming, the eval-template-reuse claim, the unverified warning-header mechanics, and the material duplicated from the Command Center guardrails page. Add the deliver-or-block mental model diagram. Signed-off-by: khushalsonawat <khushalsonawat@gmail.com>
docs(evaluation): Guardrails (concept)
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
2 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>