Repository navigation
feat(dashboard): add text search to the Experience list - #2033
Merged
Merged
Conversation
yydawx
requested review from
AlexStocks,
PsiACE,
frostming and
zhanghuidinah
as code owners
October 16, 2026 18:09
1 task
|
|
yydawx
force-pushed
the
feat/1928-exp-search
branch
from
October 19, 2026 05:57
ce2bd6e to
54cb847
Compare
hidb4ai
requested changes
Oct 19, 2026
hidb4ai
left a comment
hidb4ai
left a comment
Contributor
There was a problem hiding this comment.
Please preserve the new Experience search state through detail/back navigation and retry links; the two reproduced failures are described inline.
yydawx
force-pushed
the
feat/1928-exp-search
branch
2 times, most recently
from
October 20, 2026 07:39
230edd9 to
110e654
Compare
The Methods page lets users search Skills but only browse Experiences, even though the family-scoped Artifact search added in oceanbase#1862 covers the family. Load the Experience collection through POST /v1/scopes/{scope_id}/artifacts/experience/search when the Methods page carries a query, and adapt each hit to the exact-revision record shape the existing card, detail and paging path already uses. An empty query keeps the cursor-based browsing, and the result window follows the Skill search so a broad query cannot render hundreds of cards at once. Page links keep the kind and the query, the clear action resets the search position, and no-match and read failures stay separate with matching Chinese and English labels. Document the search entry point on the Dashboard page-API table.
yydawx
force-pushed
the
feat/1928-exp-search
branch
2 times, most recently
from
October 20, 2026 13:50
4043966 to
b579437
Compare
hireview001
reviewed
Oct 21, 2026
Return from an Experience or Skill detail page to the result page the reader came from: collection links carry a validated return_to for the Methods page, and the detail back link, its error state, revision comparison and rollback keep it. Return navigation applies the current interface language. Retry after a failed search request keeps experience_page because the same-page link allowlist includes it, while a new query still starts on the first page. Both families share the Methods template, so Skill detail pages gain the same return path.
yydawx
force-pushed
the
feat/1928-exp-search
branch
from
October 21, 2026 03:35
b579437 to
f0ba4f3
Compare
yydawx
marked this pull request as draft
October 21, 2026 03:39
Contributor
Author
|
I converted this pr to draft for now as I will test more |
Cover supported query text and its encoded size, rejected scope, path, kind and page contexts, Skill result-page restoration, and the detail read-error back path.
Assert the Skill detail back link returns to the filtered Methods page instead of only checking the kind.
Contributor
Author
|
I now add more tests and do more E2E tests. It's ready for review |
yydawx
marked this pull request as ready for review
October 21, 2026 08:31
A 404 from an out-of-range Experience or Skill result page kept the invalid page in the Retry link, so retrying could never recover. Drop result-page parameters on 404 while keeping them for transient failures.
Share detail and retry link extraction, merge the experience paging and detail-back cases, and cover the edge query strings in one parametrized case.
AlexStocks
reviewed
Oct 21, 2026
Build the oversized query inside the test and use short ids for edge query cases so PYTEST_CURRENT_TEST stays within Windows' environment limit.
Contributor
Author
Contributor
Author
1 similar comment
Contributor
Author
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
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.
Which issue or RFC does this PR close?
Closes #1928
Rationale for this change
The Methods page supports Skill search but only allows browsing Experiences, even though the family-scoped Artifact search added in #1862 already covers the family. A long Experience collection therefore cannot be narrowed, while the Skill tab can.
What changes are included in this PR?
DashboardAPI.records()loads the Experience collection throughPOST /v1/scopes/{scope_id}/artifacts/experience/searchwhen the Methods page carries a query, and adapts each hit (contentandlineage) to the exact-revision record shape the existing card, detail and paging path already renders.kindand the query; the clear action resets the search position.search_experiences,experiences_limitandexperiences_no_matchlabels in both catalogues.The page still reads through the public HTTP surface with the caller's credentials, so scope authorization, result limits and content validation are unchanged.
Are there any user-facing changes?
Yes. The Experience tab on the Dashboard Methods page gains a search form: a query returns matching approved, active Experiences in the selected scope, and clearing it restores browsing. No API, storage or migration change.
How was this change tested?
uv run pytest tests/test_dashboard.py tests/test_dashboard_profile_handoff.py tests/test_dashboard_markdown.py→ 100 passed.ruff check,ruff format --check, andty check src/powercontext/server/dashboardpassed.scripts/dashboard_browser.cjspassed against a local Server; a targeted Playwright matrix covered search/pagination, detail back, language/theme, compare/rollback and malformed return contexts.AI usage statement
Built with OpenAI Codex (GPT-5-class coding agent) in the Codex desktop app, working in a local clone of this repository. I used it to read the issue and the #1862 search API, locate the Dashboard read path (
dashboard/api.py,content.py,pagination.py,methods.html, the label catalogues), draft the implementation and the behavior tests, and run the checks locally. Every change was reviewed before submitting, and the commit contains no generated code that I did not read.Two useful corrections from the tool: its first version of the pagination test asserted a specific card order after search, which the retrieval order does not guarantee, so the test now checks the union of result pages instead. It also first assumed the "distinct messages" requirement needed a new error surface, when the existing
read_errorcomponent already separated read failures from empty results, which kept the diff small.