Sitelet https://github.com/GoogleCloudPlatform/evalbench/pull/697
Skip to content

feat(mcp_readability): render change-tracking banner and per-tool markers in HTML - #697

Open
akangsha7 wants to merge 1 commit into
evalbench/mcp-readability-scorer-carry-forwardfrom
evalbench/mcp-readability-html-provenance
Open

akangsha7 wants to merge 1 commit into
evalbench/mcp-readability-scorer-carry-forwardfrom
evalbench/mcp-readability-html-provenance

Conversation

@akangsha7

@akangsha7 akangsha7 commented Nov 2, 2026 •

Copy link
Copy Markdown
Collaborator

Second part of wiring baseline carry-forward into mcp_readability. Builds on evalbench/mcp-readability-scorer-carry-forward (#696).

Renders the carry-forward provenance summary in the HTML feedback column (mcp_readability_llm_feedback_html) so engineers can see at a glance whether a tool was carried forward or re-judged.

What changed

  • evalbench/scorers/mcp_readability/style_readability.py:
    • Add _provenance_banner() and render <p class='mcp-provenance'><b>Change tracking:</b> ...</p> above the summary in to_html().
    • Append a per-tool status marker (unchanged since the previous review, re-judged (schema changed), or new tool) to each tool's <h4> heading.

Testing

  • pytest evalbench/test/mcp_carry_forward_scorer_test.py evalbench/test/mcp_readability_test.py: 50 passed.
  • pycodestyle . --config=.pycodestyle: clean.

@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-html-provenance branch from ab7e6bb to fe24985 Compare November 2, 2026 23:30
@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-scorer-carry-forward branch 2 times, most recently from 534d5bb to f4498cd Compare November 3, 2026 04:05
@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-html-provenance branch from fe24985 to 36b3f54 Compare November 3, 2026 04:05
@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-scorer-carry-forward branch from f4498cd to d2fd215 Compare November 3, 2026 04:38
@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-html-provenance branch from 36b3f54 to 6e7ef98 Compare November 3, 2026 04:38
@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-scorer-carry-forward branch from d2fd215 to a21bdd8 Compare November 3, 2026 04:47
@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-html-provenance branch from 6e7ef98 to b4239e2 Compare November 3, 2026 04:47
@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-scorer-carry-forward branch from a21bdd8 to c2a853b Compare November 3, 2026 06:21
@akangsha7
akangsha7 force-pushed the evalbench/mcp-readability-html-provenance branch from b4239e2 to 947d4e0 Compare November 3, 2026 06:21
@akangsha7

Copy link
Copy Markdown
Collaborator Author

/gcbrun

@akangsha7 akangsha7 self-assigned this Nov 3, 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.

1 participant