On Windows, `HOME` is not a standard environment variable — the
equivalent is `USERPROFILE`. When MCP clients (e.g. Claude Code)
spawn the QMD server as a subprocess, they pass `USERPROFILE` but
not `HOME`. This causes QMD to fall back to `/tmp`, opening an
empty database instead of the user's actual index.
Fix: check `USERPROFILE` before falling back to `/tmp`.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When a proxy or firewall intercepts HuggingFace downloads, the cached
file is an HTML page instead of a GGUF model. Previously this surfaced
as an opaque "Invalid GGUF magic" error from node-llama-cpp.
Now we validate the GGUF magic bytes right after download, detect HTML
pages specifically, delete the bad file, and show an actionable error
message with workarounds (HF_ENDPOINT mirror, manual download path).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
When qmd update runs against an index created before case-preservation,
documents may exist under lowercase paths (e.g. "skill.md" for a file
actually named "SKILL.md"). Add findOrMigrateLegacyDocument() that:
- Falls back to a lowercase lookup when the canonical path is not found
- Renames the document path in-place via UPDATE OR IGNORE
- Manually rebuilds the FTS entry (FTS5 INSERT OR REPLACE does not
reliably update existing rows via triggers)
- Handles UNIQUE conflicts gracefully (returns null on conflict)
Embeddings are keyed by content hash, so the rename preserves all
existing vectors — no re-embedding required.
Both the CLI indexer and the library reindexer share the same helper,
eliminating the duplication that a previous review flagged.
Includes integration tests for: successful migration, already-lowercase
no-op, and UNIQUE conflict handling.
The blanket .toLowerCase() in handelize() drops filename casing,
which breaks path resolution on case-sensitive filesystems (Linux).
Files like README.md, CHANGELOG.md, and SKILL.md become unreachable
when the index stores them as readme.md, changelog.md, skill.md.
Since FTS5 already performs case-insensitive matching via the
unicode61 tokenizer, lowercasing the stored path provides no search
benefit — it only corrupts the metadata used to locate files on disk.
Remove .toLowerCase() and update all affected test expectations.
but that function has no callers — the CLI's outputResults() uses its
own inline JSON formatting that destructured only .snippet from
extractSnippet(), discarding .line. Extract the full SnippetResult
and spread the line field into the JSON output object.
Closes#505
Bun runs all test files in a single process, so module-level state
leaks between files. The getDefaultDbPath test now resets the
_productionMode flag before asserting it throws, fixing the flaky
failure on Bun (ubuntu-latest) in CI.
When the embed model is configured via YAML (not env var), formatDocForEmbedding
and formatQueryForEmbedding callers in store.ts would fall back to the default
model, producing the wrong prompt format. This adds a public embedModelName
getter on LlamaCpp and threads it through all five call sites.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Separate hardcoded default from env var in DEFAULT_EMBED_MODEL so the
constructor can resolve: config param > env var > hardcoded default.
Also add env var support for QMD_GENERATE_MODEL and QMD_RERANK_MODEL.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds a benchmark harness that measures search quality across backends.
Given a fixture file with queries and expected results, it runs each
query through BM25, vector, hybrid (no rerank), and full pipeline,
then reports precision@k, recall, MRR, F1, and latency.
This is primarily a regression testing tool — users create fixtures
for their own vaults to catch quality regressions after config or
index changes. Ships with an example fixture against the eval-docs
test collection to demonstrate the format.
New files:
src/bench/bench.ts — main runner
src/bench/score.ts — precision, recall, MRR, F1, path matching
src/bench/types.ts — fixture and result types
src/bench/fixtures/ — example fixture
test/bench-score.test.ts — unit tests for scoring (16 tests)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Adds QMD_LLAMA_GPU env var (set to false/off/none to force CPU) and
wraps getLlama() in try/catch so Vulkan/CUDA init failures on headless
or driverless machines fall back gracefully instead of crashing the
node process with an uncatchable C++ terminate().
- Restore .toLowerCase() in handelize (was dropped, both test files
expected it inconsistently)
- Convert dots to dashes in filename body (e.g. v2.0 -> v2-0), keeping
only the extension dot. Tobi confirmed this is the intended behavior.
- Align both test/store.test.ts and test/store.helpers.unit.test.ts to
match (they had diverged, one expected case-preserved, one lowercase)
- Adjust 'ensureVecTable recreates' test to expect throw behavior
(matches #501 dimension-mismatch fix)
- Restore .toLowerCase() in handelize (was dropped somewhere, tests expect it)
- Update dimension-mismatch test to expect throw instead of silent rebuild
(matches new behavior from #501)
- Fix one stale test expectation for preserved dots in filenames
Brace expansion patterns like `{doc1,doc2}.md` or `collection/{a,b}.md`
were incorrectly parsed as comma-separated file lists instead of being
passed to the glob matcher (picomatch). This happened because the
comma-detection heuristic only checked for `*` and `?` but not `{`.
Also adds `collection/path` matching in `matchFilesByGlob` so patterns
like `my-collection/{file1,file2}.md` work — previously the glob only
matched against `qmd://collection/path` (virtual) and `path` (relative
to collection root), missing the `collection/path` form.
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
When switching to an embedding model with different dimensions,
ensureVecTableInternal() silently drops the vector table and all
embeddings are lost. Users only discover this when semantic search
returns empty results.
Throw an error instead, telling users to run 'qmd embed -f' to
explicitly re-embed. This is safe because embed -f calls
clearAllEmbeddings() which drops the table before ensureVecTable
is reached.
Related to #497
Co-authored-by: JohnRichardEnders <john@telli.com>
Without an explicit contextSize, node-llama-cpp defaults to "auto" which
allocates the model's full training context (often 32k). For embedding
chunks that are typically ~900 tokens this wastes ~3.5 GB of KV cache
per context on Apple Silicon unified memory.
Default to 2048 (matching the rerank context pattern) and allow override
via QMD_EMBED_CONTEXT_SIZE for users with larger chunks.
Addresses #329, related to #297
Co-authored-by: JohnRichardEnders <john@telli.com>
sanitizeFTS5Term stripped all non-letter/non-number characters including
underscores, causing snake_case identifiers like `my_variable` to become
`myvariable` and silently fail BM25 matches.
Add underscore to the preserved character set in the Unicode regex.
Export the function and add unit tests covering snake_case, contractions,
punctuation stripping, and unicode.
Fixes#305
Co-authored-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The MCP status tool shows collection paths but not names,
making it impossible for agents to discover valid collection
filter values. The CLI 'qmd status' already shows names.
Add col.name prefix to each collection line in the status
tool response.
Resolve conflicts: combine AST chunking args (filepath, chunkStrategy)
with abort signal parameter from #458.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
