The broad candidate leg was replaced with an empty array in
`matchMode: "exact"`, so candidates came only from the server-side
`subject ~ query` filter and any work package whose match lived in a
description or a comment was unreachable. That inverts the parameter's
own guidance: exact mode is advertised for literal strings known
verbatim — error codes, ticket refs, commit SHAs — which are exactly the
strings that live in descriptions and comments rather than subjects.
Spec section 3.3 mandates two parallel requests with no exact-mode
carve-out.
Both legs now also page the candidate window in batches of 100 up to the
unchanged 250 cap, the way meeting search already does. Every tool shape
in this repo caps `pageSize` at 100 and OpenProject enforces a
server-side maximum, so a single `pageSize: 250` request could silently
return fewer candidates than the code claimed, with `response.total`
ignored so nothing detected it.
Also corrects `offset` on the search page from an item index to the
1-based page number its schema documents, and drops the duplicate
`items` array from `WorkPackageSearchPage` (kept on `PaginatedResult`
and the non-search list paths).
The existing `workPackageFixtureClient` returns every element for any
collection request and so cannot see a subject filter at all, which is
why this was invisible to the suite. New tests use a fixture that
honours the filter server-side; the old fixture is untouched.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Documents the fuzzy-search-by-default change shipped across tasks 1-10:
appends ADR-019 (fuzzy default, hand-rolled zero-dependency scorer, why
fuse.js/fastest-levenshtein were rejected), describes the src/search/
pure computation unit in ARCHITECTURE.md, updates README.md's tool
table/count (18 -> 19) and adds a Search behaviour section, and updates
TODO.md with the completed milestone, two accepted limitations, and
three tracked follow-ups. No source or test files were changed;
bun test (349 pass) and bun run typecheck remain unchanged.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
The "results are stable" test queried "budget", a single-match query,
so the sort had nothing to reorder and would pass even with a broken
or removed ascending-id tie-break. Switch to "muller", where docs 1
and 5 tie on score, so the assertion actually exercises the tie-break.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
The test queried 'budget', which only one corpus document matches, so it
returned a single-element array with nothing to reorder — it would pass
even with the tie-break removed. Switched to 'muller', where docs 1 and 5
tie at the same score and ordering is decided solely by the ascending-id
tie-break.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
The test used query 'certificat', which is a literal substring of the
fixture's 'certificate' comment text. Exact mode is substring containment
by design, so the query matched and the toHaveLength(0) assertion could
never pass. Replaced with an insertion typo, which cannot be a substring
of the shorter original.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Adds optional query/matchMode to listWorkPackageActivities, ranking the
already-fetched activity list client-side via rankRecords (no extra API
calls). Exposes the same fields on the
openproject_list_work_package_activities tool. When no query is supplied,
every activity is returned exactly as before, and onlyComments continues
to filter independently.
Also fixes a defect in the brief's exact-mode test fixture: the specified
typo query "certificat" is a literal prefix substring of "certificate", so
byte-for-byte exact containment matching (scoreExact in src/search/rank.ts)
finds it regardless of typo tolerance, making the test unable to pass
against a correct implementation. Swapped in "certificatee" (not a
substring of "certificate") to preserve the test's intent without touching
the assertion. Updated one pre-existing test-tools.test.ts call site to
supply matchMode, now required by the extended args type.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
The union candidate fetch wrapped both listWorkPackages() legs in an
unconditional .catch(() => ({ items: [] })), swallowing 401/403/429 and
OpenProjectAuthenticationError along with genuinely transient failures. An
expired token would silently degrade to "0 candidates", and the LLM would
report work packages as not found rather than surfacing the auth failure.
Replace the inline catches with a shared listCandidates() helper that uses
isFatalSearchError (already used by searchPipeline) to rethrow fatal errors
and only swallow non-fatal ones, preserving the property that a bad response
on one leg doesn't kill the other leg's candidates.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Registers openproject_search_work_packages (readOnly: true), taking the tool
registry from 18 to 19. searchWorkPackages now unions a precise server-side
subject filter with a broad recency-ordered candidate window, ranks via the
shared search pipeline, and deep-enriches top candidates with description and
comment text. Exact-mode requests skip the broad query entirely.
searchWorkPackages's return type changes from a flat PaginatedResult<WorkPackageSummary>
to WorkPackageSearchPage, nesting the summary under `workPackage` alongside
score, matchedFields, and an optional snippet.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Retain the page body from the formattable text field (already fetched
per-page, zero extra API calls) and rank title/text matches with
rankRecords instead of exact substring on title alone. searchWikiPages
now returns WikiPageSearchResult[] with score, snippet, and
matchedFields; empty-query behavior (project scoping + limit, no
ranking) is preserved. Adds a matchMode tool parameter defaulting to
fuzzy.
Updates the pre-existing tool-metadata assertion in tests/tools.test.ts
to match the new search_wiki_pages description introduced here.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Routes searchMeetings through the shared searchPipeline instead of a
hand-rolled Promise.all fan-out, fixing two live bugs: an expired token
on agenda-item fetch previously produced "no results" (swallowed via
.catch(() => undefined)) instead of propagating the auth error, and up
to 250 candidate meetings could fire simultaneous agenda-item requests.
The pipeline caps enrichment at 25 candidates with 8 in flight, and
isFatalSearchError now lets 401/403/429 propagate while other failures
degrade a single record and set `degraded`.
Also removes the old substring-based extractSnippet from
src/services/meetings.ts in favor of src/search/snippet.ts, and adds
matchMode ("fuzzy" | "exact") to SearchMeetingsParams and the
openproject_search_meetings tool schema.
Updated two pre-existing tests in tests/meetings.test.ts whose fixtures
assumed the old implementation's API surface: one didn't stub the
meeting-detail endpoint the pipeline's deep enrichment now also fetches
(for participants), and one assumed a 50-item candidate page size where
the new pipeline hardcodes 100. Updated one assertion in
tests/tools.test.ts that checked the old literal tool description.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Adds searchPipeline, which ranks candidates shallowly at no API cost, then
enriches a bounded subset (DEFAULT_ENRICH_LIMIT=25, DEFAULT_CONCURRENCY=8)
and re-ranks, taking max(shallow, deep) so a strong title match is never
diluted by unrelated deep content. Auth and rate-limit errors abort the
whole search via isFatalSearchError instead of degrading silently; other
enrichment failures degrade one record and are counted. A recency backfill
spends leftover enrichment budget on the most recently updated candidates
so a deep-only match is reachable at all. Adds the src/search/ barrel
export.
Reconciles the previously uncommitted pipeline.ts (written against an
earlier, defective version of this task's brief) with the corrected brief:
mapWithConcurrency's phase-2 callback is now given an explicit
EnrichedCandidate return type so its `deep: TDeep | undefined` field is a
single type rather than an inferred union, which is what the `entry is
Pair` narrowing predicate requires under TS2677.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
mapWithConcurrency's callback return was inferred as a union of two object
literal types rather than one type with deep: TDeep | undefined, so the
'entry is Pair' predicate was rejected: a type predicate's type must be
assignable to its parameter's type. Added an explicit EnrichedCandidate
annotation on the call. Tests passed but tsc did not, so this was only
visible via typecheck.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Adds boundedLevenshtein with row-minimum early exit, token scoring
(exact / prefix / contains / shared-prefix / edit distance), and
coverage-weighted field aggregation scaled by BASE_SCALE so that only a
verbatim phrase match reaches 1.0.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
The guard tests used 3-char query tokens whose field tokens were literal
continuations ('car'/'carpet'), so the startsWith rule returned 0.9 before
the shared-prefix rule was reached. With MIN_SHARED_PREFIX=4 the rule is
unreachable for any 3-char token, making the assertion impossible to
satisfy. Swapped in cases where neither token is a true prefix of the
other, which actually exercise the guard.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Replace manual exec loop with matchAll to eliminate cross-call mutable
lastIndex state on module-level TOKEN_PATTERN regex. matchAll clones
the regex internally, ensuring no shared state while maintaining
identical behavior and original-text offset semantics.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Hand-computing the plan's own test expectations against its own formula
showed four assertions could not pass:
- Body-only matches maxed at 1.0*1/3 = 0.333, below minScore 0.35, so no
low-weight field could ever return a result. The gate now uses the raw
field score; the weighted score only orders.
- The phrase bonus was dead code whenever every token matched, since the
base already saturated at 1.0. Base is now scaled by 0.85.
- Natural-language queries tanked the coverage multiplier; query tokens
are now stopword-filtered.
- 'decide' vs 'decision' is 4 edits apart and unmatchable; scoreToken
gains a shared-prefix rule.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Adds an 11-task TDD implementation plan covering the pure search unit
(tokenize, score, rank, snippet, pipeline), per-service integration for
meetings, wikis, work packages and activities, a recall acceptance
suite, and documentation.
Corrects two errors found while writing the plan:
- The spec's behaviour table named an openproject_search_work_packages
tool that does not exist. searchWorkPackages is a service helper that
only tests call, so ranked work package search would have been
unreachable by any LLM. The plan now registers the tool (18 -> 19) and
leaves openproject_list_work_packages as a pure server-side filter.
- Plan Task 6 originally identified matching agenda items by substring
searching the meeting-level snippet, which cannot work because
snippets are trimmed and ellipsized. Agenda items are now ranked
directly.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Specifies replacing exact substring matching with token-aware fuzzy
matching as the default across all search tools, plus relevance ranking
and wider content coverage (wiki body text, work package descriptions
and comments, meeting participants and project names).
Key decisions:
- Hand-rolled zero-dependency scorer in src/search/; fuse.js rejected
because its Bitap implementation caps patterns at 32 chars and targets
short fields, not the long-text matching required here.
- matchMode: 'fuzzy' | 'exact' escape hatch, defaulting to fuzzy, with
exact mode preserving current behaviour byte-for-byte.
- Two-phase pipeline: cheap shallow rank, deep-enrich top 25 with
bounded concurrency, recency backfill for deep-only matches.
Also records two known limitations left out of scope: the two-phase
recall ceiling and the wiki cache's ID 1..50 probe window.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
