Skip to content

docs: add SRE root cause accuracy evaluator page - #852

Closed
NivedhN160 wants to merge 36 commits into
future-agi:devfrom
NivedhN160:feat/sre-root-cause-accuracy-evaluator
Closed

docs: add SRE root cause accuracy evaluator page#852
NivedhN160 wants to merge 36 commits into
future-agi:devfrom
NivedhN160:feat/sre-root-cause-accuracy-evaluator

Conversation

@NivedhN160

Copy link
Copy Markdown

Adds documentation page for SRE root cause accuracy evaluator.

khushalsonawat and others added 30 commits June 25, 2026 21:03
docs(command-center): add overview video to the Agent Command Center page
docs(command-center): autoplay overview video on #watch links
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Wipe the fictional v1.0.0–v2.0.0 entries so the page starts empty, ready
for the first real automated release below the insert marker. Drop the now
unused Callout import.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Add escapeMdx() and apply it to bullet lines in mapped sections and
breaking-changes lines so <, { and } in release notes can't break the MDX
build. Add a test covering the escaping and a lockstep comment tying
SECTION_MAP to release-please-config.json.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…angelog

The sync was pointed at src/pages/changelog.mdx — orphan scaffold with no
navigation links. The real, actively maintained surface is
src/pages/docs/release-notes.mdx (weekly product-voice entries). Transform
now emits that page's exact format (styled wrapper + Features /
Bugs/Improvements / Breaking Changes subsections, Bug Fixes + Performance
merged into the existing Bugs/Improvements bucket) and inserts new versions
on top of the existing entries. Orphan changelog.mdx deleted. Tests: 5/5.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Owner decision: this PR no longer touches the changelog page at all —
neither deletion nor placeholder wipe. It is restored byte-identical to
main. The release-notes sync is unaffected (targets docs/release-notes.mdx).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
ci(changelog): sync docs changelog from platform releases
…UB_TOKEN

The org forbids GitHub Actions from creating PRs with GITHUB_TOKEN; App
installation tokens are exempt (release bot App already installed on this
repo). Mint a scoped token and use it for both the branch push and
gh pr create — consistent with how future-agi opens its release PRs, and
avoids relaxing the org-wide setting.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…c-app-token

ci(changelog): open release-notes PR with App token (not GITHUB_TOKEN)
@khushalsonawat
khushalsonawat self-requested a review September 8, 2026 15:43
@khushalsonawat
khushalsonawat changed the base branch from main to dev September 8, 2026 15:44
@khushalsonawat

khushalsonawat commented Sep 8, 2026

Copy link
Copy Markdown
Member

Hello @NivedhN160,
Thank you for contributing to Future AGI docs. This PR adds the documentation for a Root cause accuracy evaluator under the built-in eval section. However, we currently don't have a built-in Root cause evaluator. Do you happen to have a PR open in the monorepo or are currently in discussion with the internal team to add one?

@NivedhN160

Copy link
Copy Markdown
Author

Thanks @khushalsonawat — you're right. There is no Root Cause Accuracy evaluator in the monorepo yet.

I don't have a monorepo PR open for it. Happy to open one that adds SRERootCauseAccuracy under futureagi/agentic_eval/core_evals/fi_evals/ and register it in eval_type.py, then keep this docs PR in sync (or close this and reopen after the evaluator merges).

Should I open the monorepo issue/PR first, or is there an internal design you want aligned before implementation?

@khushalsonawat

khushalsonawat commented Sep 8, 2026

Copy link
Copy Markdown
Member

@NivedhN160,
The built-in evals are the ones that the internal team builds and posts them as common evals templates to be used for everyone. And that is what's present in the built-in evals list. If none of the built-in evals seem to fit your use case, you can create a custom eval that fits your use-case.The guide to create one is here

@NivedhN160

Copy link
Copy Markdown
Author

Understood — thanks for clarifying. Built-in templates are internal-only.

I'll close this PR. For the root-cause use case I'll use a custom eval via the guide you linked instead of proposing a new built-in.

@NivedhN160 NivedhN160 closed this Sep 8, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants