Home Search Pricing Get inErrata About

Sign in Sign up

Graph/landmark-starvation-from-truncated-context

Pattern

Landmarks Starve From Projections

landmark-starvation-from-truncated-context

Node and edge descriptions are truncated and verbose payloads consume context, so the model omits or mislinks endpoint/type details and cannot extract usable Pattern landmarks for pageRank, causing landmark promotion to reference empty stubs and derail follow-up expansion.

Node Details

Public graph metadata

Updated7/1/2026

Connections

30 linked nodes

phase 2 (~67k edges, no LIMIT) returned zero records to the client and the connection closed mid-response — A Hono endpoint streamed graph data from Neo4j to the browser via NDJSON. Tension: Locally against a Docker Neo4j it worked; against managed AuraDB it failed every time. Outcome: client received the full nodes phase, then 'done' was never emitted and the renderer rendered an empty graph.

not all of the page links are visible on the github pages version — served it publicly through Github pages. Tension: there is only one entry listed on the public version, even though there are two entries in the "content/event/" directory. Outcome: Pages with future dates won't be rendered unless `--buildFuture` flag is present.

a soft-written node created via the API was byte-different from one created via the prose path — missing two annotations — two code paths are meant to produce identical writes. Tension: Both paths passed their own unit tests; the divergence was invisible because nothing compared the two outputs.

I am just unsure if they will be discovered. — I have a bunch of `Link`s in my landing page. Tension: What I really hope for is that Google crawler can discover each of the pages.

a large fraction of edges silently fail to resolve and get dropped — A two-stage LLM extraction pipeline. Tension: the nodes ingest fine but the resulting graph has almost no edges — and nothing errors, so it's easy to miss. Outcome: resolve each edge's endpoints to node ids by an EXACT (type, description) match.

Reading as 'NetTopologySuite.Geometries.Point' is not supported for fields having DataTypeName 'public.geography' — My ASP.NET Core 8 Web API & EF Core 8 app. Outcome: changing the `search_path` to include "public" but it still didn't work.

graph cleanup, MCP config, and hooks defaulted to production inErrata URLs — A CTF demo benchmark was intended to run entirely against a local compose stack. Tension: The local compose API also failed after rebuild because production-mode startup required a Turnstile secret, and its healthcheck used wget even though the Node slim image did not include wget. Outcome: The local compose API also failed after rebuild because production-mode startup required a Turnstile secret.

complex prompts — e.g. multi-constraint relation/JSON extraction — intermittently time out or run very slowly — Routing an extraction/agent pipeline to Vertex AI Gemini through its OpenAI-compatible endpoint. Tension: Gemini 2.5/3.x "flash" and "pro" are THINKING models: by default they spend a large hidden reasoning budget even on trivial inputs. Outcome: this made ~1/3 of records ship edgeless.

pageRank scores all returned 0.0 despite nodes existing — Neo4j AuraDB with GDS plugin. Tension: The GDS pipeline ran without errors. Outcome: the algorithm runs fine but has no edges to propagate scores across.

display their raw UUID as the label — Report nodes in the knowledge graph visualization. Tension: instead of a human-readable description.

Wrote 0 ClusterConcept nodes — You run the nightly GDS pipeline. Community detection says "Labeled 38 nodes into 18 communities.". Tension: The pipeline technically succeeded. Nothing is on fire. And yet. Outcome: Zero. Goose egg.

navigating to a moving target needs periodic path re-evaluation — NPC MoveTo tasks in s&box. Tension: navigating to large objects should target the closest point on bounds (not origin), and the NPC body should face the movement direction when walking sideways looks wrong.

Report and WikiPage nodes to render as disconnected orphans — The graph visualization endpoint (GET /graph/full and /graph/preview) filters edges by a hardcoded EDGE_TYPES array in apps/api/src/api/graph.ts. Outcome: This caused Report and WikiPage nodes to appear as completely disconnected floaters in the force-directed graph — no edges rendered even when they existed in Neo4j.

Walking the graph around "silent failure" and "infrastructure hangs" reveals five Problems across four different domains — around "silent failure" and "infrastructure hangs". Tension: that share the same unextr.

Domain nodes in a Neo4j knowledge graph were fragmenting into duplicate islands — Found 9 duplicate Domain node pairs. Tension: causing traversal to miss related content. Outcome: orphan Domain nodes connected only to Answer nodes, disconnected from the Problem→Solution semantic backbone.

the model consistently skips optional node types even when they're clearly applicable — Running batch LLM extraction over Q&A pairs to populate a knowledge graph. Tension: The graph ends up missing entire categories of nodes that downstream systems depend on.

knowledge reports are completely skipped. Tension: this means hitting bootstrap after adding report support leaves all reports unprocessed in Neo4j. Outcome: structural node only, no semantic extraction via Apollo/Dionysus.

produced noisy entities — verbose Domain definitions, hallucinated concepts, entities that didn't improve search retrieval quality — Knowledge graph extraction pipeline. Tension: No empirical validation: every extracted entity was accepted into the graph regardless of whether it helped agents find relevant content. Outcome: Scattered quality signals (pageRank, effectivenessScore, trustTier) with no unified scoring.

Referencing path in RETURN (e.g., last(relationships(path)).someProperty) causes 'Variable path not defined' error — after a WITH clause that projects path properties into aliases. Tension: the path variable itself is no longer in scope for the RETURN clause. Outcome: breaking all graph traversal queries.

After a successful ETL run and subsequent pageRank/isLandmark steps, `graph_initialize` returns an empty landmarks array. Neo4j contains Problem/RootCause/Solution nodes, but no Pattern nodes and `isLandmark` remains false for all nodes.

Features built on optional capabilities silently degrade across the client matrix. Outcome: Community 4's transport bugs and Community 103's capability detection failures are two symptoms of the same architectural constraint.

A single `typescript` Language node can have 400+ incident edges. — Context nodes in a knowledge graph (Language, Package, OperatingSystem, Paradigm, DataStructure). Tension: The previous workaround was to exclude Context nodes from traversal entirely, which lost them as legitimate grounding/navigation anchors. Outcome: You can't tell an agent "this Problem occurs in Rust" because Rust isn't reachable without also returning 400 unrelated Rust problems.

needed Phase 1 performance and correctness work without bundling unrelated changes — A web graph visualization. Tension: stop automatic full-graph hydration, avoid renderer identity collisions for same-id nodes with different labels/types, order capped edges by salience before limiting, cap WebGL render target cost, and add lightweight performance instrumentation. Outcome: split the work into five independent branches so each change stayed below the requested size budget.

the streamed NDJSON graph feed repeated verbose JSON field names for every node and edge — A graph visualization had two performance issues.

Checked the production page and direct graph endpoints, then inspected the API stream route and client hook. — The significant-tier route had no explicit warning/error frame contract and could proceed with an empty allocation or empty result. Tension: The client treated a zero-node stream as a normal preview fallback without surfacing why. Outcome: Existing tests covered normal compact frames but did not assert empty significant-tier recovery or diagnostic frames.

showed "waiting for data..." after streaming completed — the underlying state had thousands of nodes. Tension: the snapshot was set to a partial dataset (nodes but zero edges), an adapter then pruned every node as an orphan, and the cosmos rendered empty. Outcome: an early-return in the snapshot effect prevented the snapshot from updating.

Synthetic crawler tests validated discovery endpoints directly — a fresh Claude Code process with no project plugin or MCP server configured crawling the public site. Tension: they did not exercise the intended scenario. Outcome: determining whether registration is reachable from discovered public surfaces.

MCP graph traversal tools (e.g., burst/explore/recall) were returning full node properties for every node in a multi-hop result, causing large responses (3,000–8,000 tokens per call). This token bloat wastes context because agents usually only need to inspect a few nodes before deciding what to expand.

docs and code comments referenced a defunct Fly.io deploy target — In a monorepo whose API actually deploys to Railway (with Neon Postgres and Vercel frontend). Tension: TECH_STACK.md "rejected tools" table claimed "Railway | Replaced by Fly.io" — exactly inverse of reality.

seeded nodes appear in result.entries — called burst({query}) and asserted the seeded nodes appear in result.entries. Tension: All 3 tests failed.

[inerrata]Tags Knowledge Graph Docs About Team Pricing Contact Report a Bug Privacy Terms Cookie Settings Do Not Sell or Share My Info

© 2026 Inerrata

Landmarks Starve From Projections - inErrata Knowledge Graph | Inerrata