mirror of
https://github.com/Kaelio/ktx.git
synced 2026-06-13 08:15:14 +02:00
docs(concepts): add Wiki retrieval pillar page
Adds a dedicated concept page covering the wiki side of the context layer: the page contract, the hybrid retrieval pipeline (lexical, semantic, token lanes fused by RRF), the refs/sl_refs/[[wikilink]] graph, validation that keeps edges live, and where ingest sources pages. Wired into concepts nav and cross-linked from the-context-layer to mirror the existing Semantic querying link.
This commit is contained in:
Andrey Avtomonov
parent
2f647d5c68
commit
d781a885b1
3 changed files with 285 additions and 1 deletions
|
|
@ -1,5 +1,5 @@
|
|||
{
|
||||
"title": "Concepts",
|
||||
"defaultOpen": true,
|
||||
"pages": ["the-context-layer", "semantic-layer-internals", "context-as-code"]
|
||||
"pages": ["the-context-layer", "semantic-layer-internals", "wiki-retrieval", "context-as-code"]
|
||||
}
|
||||
|
|
|
|||
|
|
@ -195,6 +195,10 @@ wiki pages are written and prunes `sl_refs` during ingest when their target
|
|||
sources are deleted or their measures are renamed - so a stale page can never
|
||||
quietly route an agent to a definition that no longer exists.
|
||||
|
||||
For how the hybrid search pipeline ranks pages, how `[[wikilinks]]` extend
|
||||
the graph, and how ingest authors pages from evidence, read
|
||||
[Wiki retrieval](/docs/concepts/wiki-retrieval).
|
||||
|
||||
The split between the two pillars is sharp:
|
||||
|
||||
| Put it in YAML | Put it in Markdown |
|
||||
|
|
|
|||
280
docs-site/content/docs/concepts/wiki-retrieval.mdx
Normal file
280
docs-site/content/docs/concepts/wiki-retrieval.mdx
Normal file
|
|
@ -0,0 +1,280 @@
|
|||
---
|
||||
title: Wiki retrieval
|
||||
description: How ktx ranks wiki pages with hybrid search, links them into a graph, and keeps both sides anchored to evidence.
|
||||
---
|
||||
|
||||
The wiki is the prose half of the context layer. Agents reach it two ways:
|
||||
they search for a page, then follow references inside the pages they
|
||||
already opened. This page covers how both work.
|
||||
|
||||
- The wiki page contract that retrieval and validation depend on.
|
||||
- The hybrid search pipeline that turns a question into ranked pages.
|
||||
- The reference graph agents traverse without rerunning search.
|
||||
- How pages get authored from evidence, and how broken edges get pruned.
|
||||
|
||||
## The wiki page contract
|
||||
|
||||
A wiki page is a Markdown file with a YAML frontmatter block. Frontmatter
|
||||
carries metadata; the prose below it is free-form. Keys are flat tokens
|
||||
(`revenue`, `mart_account_segments`), not paths, so every page is
|
||||
addressable as `[[key]]` from any other page.
|
||||
|
||||
```markdown
|
||||
# wiki/global/revenue.md
|
||||
---
|
||||
summary: Paid order value after refunds
|
||||
tags: [finance, orders]
|
||||
sl_refs: [warehouse.orders]
|
||||
refs: [segment-classification]
|
||||
usage_mode: auto
|
||||
---
|
||||
|
||||
Revenue is paid order amount after refund adjustments.
|
||||
|
||||
Use `orders.total_revenue` for recognized order value and
|
||||
`orders.order_count` for paid order volume.
|
||||
```
|
||||
|
||||
| Field | Purpose |
|
||||
|-------|---------|
|
||||
| `summary` | One-line description shown in search results and the agent's knowledge index |
|
||||
| `tags` | Topic labels mixed into the search text and used for filtering |
|
||||
| `refs` | Outgoing edges to other wiki pages by key |
|
||||
| `sl_refs` | Outgoing edges to semantic-layer sources by `connection.source` name |
|
||||
| `usage_mode` | `always`, `auto`, or `never` - whether the agent must, may, or must not surface this page |
|
||||
| `source` | Where the page came from when authored by ingest (e.g. `historic-sql`, `dbt`) |
|
||||
| `usage` | Stats attached to historic-SQL pattern pages: executions, distinct users, runtime percentiles, error rate |
|
||||
|
||||
Pages live under two scopes. `wiki/global/*.md` is the team's shared
|
||||
context; `wiki/user/<user>/*.md` is per-agent scratch space that shadows
|
||||
global pages with the same key.
|
||||
|
||||
## What retrieval does
|
||||
|
||||
A wiki search runs the same ordered steps every time.
|
||||
|
||||
1. **Normalize the query.** Lowercase, tokenize, deduplicate terms.
|
||||
2. **Score in three lanes.** Lexical (SQLite FTS5 bm25), semantic
|
||||
(cosine similarity over embeddings), and token (term-overlap fallback)
|
||||
each rank every page independently.
|
||||
3. **Fuse with Reciprocal Rank Fusion.** Each lane contributes
|
||||
`weight / (60 + rank)` to a candidate's score. Lanes that fail or
|
||||
skip are dropped, not zeroed.
|
||||
4. **Order and trim.** Sort by fused score, then by how many lanes
|
||||
matched, then by id for stable tie-breaks. Return the top `limit`
|
||||
results with their summaries.
|
||||
5. **Hydrate on demand.** The agent calls `wiki_read` to load full
|
||||
bodies for the few pages that look relevant.
|
||||
|
||||
<figure
|
||||
className="not-prose my-10 overflow-hidden rounded-lg border border-fd-border bg-fd-card shadow-sm"
|
||||
aria-label="Three retrieval lanes fused with Reciprocal Rank Fusion"
|
||||
>
|
||||
<div className="border-b border-fd-border bg-fd-muted/35 px-5 py-4">
|
||||
<p className="text-[11px] font-semibold uppercase tracking-[0.08em] text-fd-primary">
|
||||
{"Hybrid retrieval"}
|
||||
</p>
|
||||
<h3
|
||||
className="mt-1 text-base font-semibold tracking-normal text-fd-foreground sm:text-lg"
|
||||
style={{ fontFamily: "var(--font-display)" }}
|
||||
>
|
||||
{"Three lanes, one ranking"}
|
||||
</h3>
|
||||
</div>
|
||||
|
||||
<div className="p-6">
|
||||
<div className="grid gap-3 md:grid-cols-3">
|
||||
<div className="rounded-md border border-fd-border bg-fd-background p-4">
|
||||
<p className="text-sm font-semibold text-fd-foreground">{"lexical"}</p>
|
||||
<p className="mt-1 font-mono text-[11px] text-fd-muted-foreground">{"sqlite fts5 / bm25"}</p>
|
||||
<p className="mt-3 text-xs leading-5 text-fd-muted-foreground">
|
||||
{"Matches stems and phrases. Strong on the exact terms the team already uses."}
|
||||
</p>
|
||||
<p className="mt-3 text-[11px] uppercase tracking-[0.08em] text-fd-muted-foreground">
|
||||
<span className="text-fd-foreground">{"weight "}</span>{"1.5"}
|
||||
</p>
|
||||
</div>
|
||||
|
||||
<div className="rounded-md border border-fd-border bg-fd-background p-4">
|
||||
<p className="text-sm font-semibold text-fd-foreground">{"semantic"}</p>
|
||||
<p className="mt-1 font-mono text-[11px] text-fd-muted-foreground">{"cosine over embeddings"}</p>
|
||||
<p className="mt-3 text-xs leading-5 text-fd-muted-foreground">
|
||||
{"Catches synonyms and paraphrases the lexical lane misses."}
|
||||
</p>
|
||||
<p className="mt-3 text-[11px] uppercase tracking-[0.08em] text-fd-muted-foreground">
|
||||
<span className="text-fd-foreground">{"weight "}</span>{"2"}
|
||||
</p>
|
||||
</div>
|
||||
|
||||
<div className="rounded-md border border-fd-border bg-fd-background p-4">
|
||||
<p className="text-sm font-semibold text-fd-foreground">{"token"}</p>
|
||||
<p className="mt-1 font-mono text-[11px] text-fd-muted-foreground">{"term-overlap fallback"}</p>
|
||||
<p className="mt-3 text-xs leading-5 text-fd-muted-foreground">
|
||||
{"Always available, so short queries still produce candidates."}
|
||||
</p>
|
||||
<p className="mt-3 text-[11px] uppercase tracking-[0.08em] text-fd-muted-foreground">
|
||||
<span className="text-fd-foreground">{"weight "}</span>{"0.75"}
|
||||
</p>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<div className="mt-5 rounded-md border border-fd-primary/40 bg-fd-background p-4">
|
||||
<p className="text-[11px] font-semibold uppercase tracking-[0.08em] text-fd-primary">
|
||||
{"Reciprocal Rank Fusion"}
|
||||
</p>
|
||||
<p className="mt-1 font-mono text-[12px] text-fd-foreground">
|
||||
{"score = Σ weight / (60 + rank)"}
|
||||
</p>
|
||||
<p className="mt-2 text-xs leading-5 text-fd-muted-foreground">
|
||||
{"Pages that rank well in multiple lanes outscore pages that rank well in only one."}
|
||||
</p>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<figcaption className="border-t border-fd-border bg-fd-muted/25 px-5 py-3 text-[11.5px] leading-5 text-fd-muted-foreground">
|
||||
<span className="font-medium text-fd-foreground">{"Defaults are tunable. "}</span>
|
||||
{"Lane weights and the RRF constant K are configuration, not assumptions."}
|
||||
</figcaption>
|
||||
</figure>
|
||||
|
||||
The text each lane scores is built deterministically: page key, summary,
|
||||
body, and tags concatenated in that order. A precise summary and the
|
||||
right tags make a page reachable before its body matches anything.
|
||||
|
||||
## The page graph
|
||||
|
||||
Two frontmatter fields and one inline syntax turn the wiki into a graph
|
||||
the agent traverses without re-running search.
|
||||
|
||||
| Edge | Source | Target |
|
||||
|------|--------|--------|
|
||||
| `sl_refs: [warehouse.orders]` | Frontmatter | Semantic source by name |
|
||||
| `refs: [segment-classification]` | Frontmatter | Another wiki page by key |
|
||||
| `[[segment-classification]]` | Inline in body | Another wiki page by key |
|
||||
|
||||
`refs` stays in the prose layer; `sl_refs` crosses into the executable
|
||||
half of the context layer. Inline `[[wikilinks]]` are extracted from
|
||||
page bodies at validation time and treated as declared `refs`.
|
||||
|
||||
<figure
|
||||
className="not-prose my-10 overflow-hidden rounded-lg border border-fd-border bg-fd-card shadow-sm"
|
||||
aria-label="Example wiki cross-reference graph"
|
||||
>
|
||||
<div className="border-b border-fd-border bg-fd-muted/35 px-5 py-4">
|
||||
<p className="text-[11px] font-semibold uppercase tracking-[0.08em] text-fd-primary">
|
||||
{"Anatomy of a traversal"}
|
||||
</p>
|
||||
<h3
|
||||
className="mt-1 text-base font-semibold tracking-normal text-fd-foreground sm:text-lg"
|
||||
style={{ fontFamily: "var(--font-display)" }}
|
||||
>
|
||||
{"Edges to prose, edges to SQL"}
|
||||
</h3>
|
||||
</div>
|
||||
|
||||
<div className="p-6">
|
||||
<div className="grid gap-3 md:grid-cols-[1fr_1fr]">
|
||||
<div className="rounded-md border-2 bg-fd-background p-4" style={{ borderColor: "#10b981" }}>
|
||||
<p className="font-mono text-[11px] font-semibold" style={{ color: "#10b981" }}>
|
||||
{"wiki/global/revenue.md"}
|
||||
</p>
|
||||
<p className="mt-2 text-sm font-semibold text-fd-foreground">{"revenue"}</p>
|
||||
<p className="mt-3 text-[11px] uppercase tracking-[0.08em] text-fd-muted-foreground">
|
||||
{"declares"}
|
||||
</p>
|
||||
<ul className="mt-1 space-y-1 text-xs text-fd-muted-foreground">
|
||||
<li><span className="font-mono text-fd-foreground">{"sl_refs"}</span>: warehouse.orders</li>
|
||||
<li><span className="font-mono text-fd-foreground">{"refs"}</span>: segment-classification</li>
|
||||
</ul>
|
||||
</div>
|
||||
<div className="rounded-md border bg-fd-background p-4" style={{ borderColor: "#10b981", borderStyle: "dashed" }}>
|
||||
<p className="font-mono text-[11px] font-semibold" style={{ color: "#10b981" }}>
|
||||
{"wiki/global/segment-classification.md"}
|
||||
</p>
|
||||
<p className="mt-2 text-sm font-semibold text-fd-foreground">{"segment-classification"}</p>
|
||||
<p className="mt-3 text-[11px] uppercase tracking-[0.08em] text-fd-muted-foreground">
|
||||
{"declares"}
|
||||
</p>
|
||||
<ul className="mt-1 space-y-1 text-xs text-fd-muted-foreground">
|
||||
<li><span className="font-mono text-fd-foreground">{"sl_refs"}</span>: warehouse.customers</li>
|
||||
</ul>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<div className="my-4 grid gap-2 text-center text-xs font-medium text-fd-muted-foreground md:grid-cols-[1fr_1fr]">
|
||||
<div>{"revenue → warehouse.orders · sl_refs"}</div>
|
||||
<div>{"revenue → segment-classification · refs"}</div>
|
||||
</div>
|
||||
|
||||
<div className="grid gap-3 md:grid-cols-[1fr_1fr]">
|
||||
<div className="rounded-md border-2 bg-fd-background p-4" style={{ borderColor: "#3b82f6" }}>
|
||||
<p className="font-mono text-[11px] font-semibold" style={{ color: "#3b82f6" }}>
|
||||
{"semantic-layer/warehouse/orders.yaml"}
|
||||
</p>
|
||||
<p className="mt-2 text-sm font-semibold text-fd-foreground">{"warehouse.orders"}</p>
|
||||
<p className="mt-1 text-xs text-fd-muted-foreground">{"grain: order_id · measure: total_revenue"}</p>
|
||||
</div>
|
||||
<div className="rounded-md border-2 bg-fd-background p-4" style={{ borderColor: "#3b82f6" }}>
|
||||
<p className="font-mono text-[11px] font-semibold" style={{ color: "#3b82f6" }}>
|
||||
{"semantic-layer/warehouse/customers.yaml"}
|
||||
</p>
|
||||
<p className="mt-2 text-sm font-semibold text-fd-foreground">{"warehouse.customers"}</p>
|
||||
<p className="mt-1 text-xs text-fd-muted-foreground">{"grain: customer_id · dim: segment"}</p>
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<figcaption className="border-t border-fd-border bg-fd-muted/25 px-5 py-3 text-[11.5px] leading-5 text-fd-muted-foreground">
|
||||
{"Green nodes are wiki pages; blue nodes are semantic sources."}
|
||||
</figcaption>
|
||||
</figure>
|
||||
|
||||
## Keeping the graph live
|
||||
|
||||
A page that references a deleted source is worse than no reference at
|
||||
all - it sends the agent confidently to a definition that no longer
|
||||
exists. **ktx** prevents that with three layered checks:
|
||||
|
||||
- **At write time.** Every `refs` entry and `[[wikilink]]` is validated
|
||||
against the pages visible in the current scope. A write that targets
|
||||
a missing page is rejected before any file changes.
|
||||
- **At ingest time.** Adapters prune `sl_refs` when the target source
|
||||
is deleted, mark stale pattern pages with `stale_since`, and set
|
||||
`archived_since` on retired pages instead of removing them silently.
|
||||
- **At session end.** Every page touched by an ingest run is re-scanned
|
||||
for references that resolved at write time but no longer point at
|
||||
a live target. Dangling pairs are reported so the next iteration can
|
||||
fix them.
|
||||
|
||||
## Where the pages come from
|
||||
|
||||
**ktx** writes wiki pages from evidence, not free invention. Each input
|
||||
contributes a different kind of page, and accepted edits feed the next
|
||||
ingest as input.
|
||||
|
||||
| Evidence | What it produces |
|
||||
|----------|------------------|
|
||||
| Schema scans | One page per material table, with grain, columns, and known constraints |
|
||||
| Query history | Pattern pages with `usage` frontmatter for executions, distinct users, runtime percentiles, and error rate |
|
||||
| dbt manifests | Pages per model, exposure, and test, with `sl_refs` to the matching semantic source |
|
||||
| MetricFlow, Looker, Metabase | Pages per metric, explore, or saved question, linked back to the source artifact |
|
||||
| Notion, docs, analyst notes | Pages preserving business definitions, policies, and incident write-ups |
|
||||
| Agent and analyst edits | First-class input to the next ingest, not a fork |
|
||||
|
||||
Provenance stays with the page. Ingested pages keep HTML comments like
|
||||
`<!-- from: raw-sources/.../cards/69.json -->` inline, so a reviewer can
|
||||
walk from the prose back to the artifact that produced it.
|
||||
|
||||
## Agent usage notes
|
||||
|
||||
Point an agent at this page when it needs to explain why a wiki search
|
||||
returned the pages it did, why a write was rejected, or how the wiki
|
||||
stays in step with the semantic layer.
|
||||
|
||||
| Agent task | Relevant section | Next page |
|
||||
|------------|------------------|-----------|
|
||||
| Explain why two searches return different pages for the same query | What retrieval does | [ktx wiki](/docs/cli-reference/ktx-wiki) |
|
||||
| Decide whether to add a `refs` or `sl_refs` entry | The page graph | [Writing Context](/docs/guides/writing-context) |
|
||||
| Repair a wiki write rejected for missing references | Keeping the graph live | [Writing Context](/docs/guides/writing-context) |
|
||||
| Describe how historic SQL becomes a wiki page | Where the pages come from | [Building Context](/docs/guides/building-context) |
|
||||
| Explain raw-source provenance comments | Where the pages come from | [Context as Code](/docs/concepts/context-as-code) |
|
||||
Loading…
Add table
Add a link
Reference in a new issue