Searching

centinel search "stormwater drainage fee"
centinel search "budget" --source tampa -n 20
centinel search "lobbyist" --snippet-chars 0        # whole chunk, not an excerpt

Flag	Default	Meaning
`-n`, `--limit`	10	maximum results
`--source`	all	restrict to one source
`--snippet-chars`	400	characters of the passage to return; `0` returns all of it

What comes back

One ranked passage per result, with everything needed to cite it:

Field	What it is
`text`	the passage
`title`	the document’s own name
`heading`	the markdown heading trail the passage sits under
`source`, `url`	which corpus, which address
`observed_at`	when we fetched it
`tool`	which extraction pipeline produced this text
`blob_sha`	hash of the original bytes as served — the evidentiary anchor
`derived_sha`	hash of the derived text the character span indexes into
`chunk_hash`	hash of the passage itself
`char_start`, `char_end`	the span inside the derived text
`also_at`	other addresses carrying this identical passage

also_at is not decoration. The same paragraph appears on fifty pages of a council site, and each of those addresses is its own document with its own bytes and its own history. Each entry carries its own hash, so you can open any of them.

Two arms, and why the second one matters

BM25 catches exact tokens. Names, motions, ordinance numbers, dollar figures — most of what people actually search meeting records for.

The vector arm closes the vocabulary gap. Measured on a real corpus: "drinking water sampling results" returns nothing from keyword search, because the water report says PWSName, Analyte and UCMR 5, and the only chunk containing the word “drinking” is a tax table about Drinking Places (Alcoholic Beverages). BM25 is behaving correctly and is still useless.

Both run, both return their top 100, the two rankings are fused, and a cross-encoder reranks the survivors. There is no flag to skip the reranker, because the measured gap is too large to make it an option. Search has the mechanism.

Read the header line

stormwater drainage fee    2 results · bm25→rerank · 397,830 chunks indexed
! keyword search only — no vectors at ~/.centinel/vectors.lance — run `centinel embed` first

The method field names which stages actually ran, and it is assembled from what ran rather than written out per call site. Four values are possible:

`method`	What it means
`bm25`	keyword only, unreranked — the weakest answer this tool gives
`bm25→rerank`	keyword, reranked — no vector table yet
`bm25+vector→rrf`	both arms fused, unreranked — reranker weights missing
`bm25+vector→rrf→rerank`	everything ran

This is the field to read first. A rank is a position inside a set and says nothing about the size of that set. The fusion weights by rank alone, so the vector arm’s rank 1 counts exactly the same whether it was drawn from 397,830 vectors or from 2,309. A partly embedded corpus therefore does not degrade gently — it promotes confident results from a tiny pool and looks identical to a complete one.

So the report always carries total_chunks_indexed beside vectors_indexed, and the terminal prints the share whenever it is not 100%. no_vectors and no_rerank carry why a stage did not run. An absent stage is a different answer, not a slower one.

What a search cannot tell you

A chunk’s absence from an arm is a fact about what has been processed, never about whether it answers your question. The same holds one stage earlier: a PDF that failed extraction is not in the index at all, and no search will report its absence. centinel list and the run report are where coverage lives.

If you searched for something you are confident is in the corpus and got nothing, the order to check is:

Was it collected? centinel list shows resource counts and liveness per source.
Was text derived from it? The run report counts unreadable documents per stage.
Was it indexed? total_chunks_indexed in the search report.
Was it embedded? vectors_indexed in the same report.

Cost

A warm process answers in about a second — the reranker dominates. A cold CLI invocation pays the model load every time: 11 seconds measured on a corpus with no vector table, so with only the 0.6B reranker loaded. A query that also builds the 4B embedder pays more.

centinel serve and centinel mcp load both once and keep them. If you are going to ask more than a couple of questions, ask them through one of those.

Next: Reading a result.