fix: review findings from Lefevre live smoke

Two small fixes informed by the 2026-05-18 live OpenRouter chapter-I run.

1. extract-entities templates (trading-literature and general-knowledge):
   the # Entity Title placeholder was interpreted by gpt-4o-mini as a
   literal heading prefix, so every entity came back as "# Entity Title:
   Bucket Shop" etc. The instruction now spells the placeholder out
   with concrete examples and an explicit "not the literal string"
   note, so smaller models hit the intended shape.

2. generate plan grows --model <id>. When supplied, the cost estimate
   pulls per-prompt and per-completion rates from the bundled
   model_rates.yaml instead of multiplying a single blended
   --cost-per-1k value across all tokens. The summary now also returns
   a separate estimated_completion_tokens field plus a cost_source tag
   ("rate_table:<model>" | "cost_per_1k_blended" | None).

This is a stopgap. LLM-WP-0005 (proposed in llm-connect this round)
will move the rate registry and token-shape problem classes upstream
so consumers stop re-implementing them.

The live smoke ran 28k prompt tokens / 7.5k completion / $0.0088
actual. With --model openai/gpt-4o-mini the plan estimate now lands at
$0.0076 (within 14% of actual) versus the prior $8.40 estimate at
--cost-per-1k 0.30.

181 tests pass, 2 skipped (both live OpenRouter smokes).

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

src/infospace_bench/profiles/general-knowledge/templates/extract-entities.md

@@ -3,9 +3,11 @@
 Profile: {{ macros.profile }}
 
 Extract reusable infospace entities from the source chunk. Return one Markdown
-bundle where each entity starts with `# Entity Title` and contains at least a
-`## Definition` section. Prefer durable concepts, claims, named methods,
-people, places, works, and objects over sentence fragments.
+bundle where each entity starts with a level-1 heading that is the entity's
+own name (e.g. `# Knowledge Artifact`, `# Source Claim` — **not** the literal
+string "Entity Title"). Each entity contains at least a `## Definition`
+section. Prefer durable concepts, claims, named methods, people, places,
+works, and objects over sentence fragments.
 
 Source title: {{ input.title }}
 Source artifact: {{ input.artifact_id }}

src/infospace_bench/profiles/trading-literature/templates/extract-entities.md

@@ -3,8 +3,10 @@
 Profile: {{ macros.profile }}
 
 Extract reusable infospace entities from the source chunk. Return one
-Markdown bundle where each entity starts with `# Entity Title` and has a
-`## Definition` section, plus a `## Category` line drawn from the list
+Markdown bundle where each entity starts with a level-1 heading that is
+the entity's name (e.g. `# Bucket Shop`, `# Tape Reading`, `# Larry
+Livingston` — **not** the literal string "Entity Title"). Each entity has
+a `## Definition` section and a `## Category` line drawn from the list
 below. Add `## Context` and `## Source Evidence` when the chunk gives
 enough material; leave them out rather than inventing detail.