mirror of
https://github.com/koala73/worldmonitor.git
synced 2026-04-26 01:24:59 +02:00
dcf73385ca
* fix(scoring): rebalance formula weights severity 55%, corroboration 15%
PR A of the scoring recalibration plan (docs/plans/2026-04-17-002).
The v2 shadow-log recalibration (690 items, Pearson 0.413) showed the
formula compresses scores into a narrow 30-70 range, making the 85
critical gate unreachable and the 65 high gate marginal. Root cause:
corroboration at 30% weight penalizes breaking single-source news
(the most important alerts) while severity at 40% doesn't separate
critical from high enough.
Weight change:
BEFORE: severity 0.40 + sourceTier 0.20 + corroboration 0.30 + recency 0.10
AFTER: severity 0.55 + sourceTier 0.20 + corroboration 0.15 + recency 0.10
Expected effect: critical/tier1/fresh rises from 76 to 88 (clears 85
gate). critical/tier2/fresh rises from 71 to 83 (recommend lowering
critical gate to 80 at activation time). high/tier2/fresh rises from
61 to 69 (clears 65 gate). The HIGH-CRITICAL gap widens from 10 to
14 points for same-tier items.
Also:
- Bumps shadow-log key from v2 to v3 for a clean recalibration dataset
(v2 had old-weight scores that would contaminate the 48h soak)
- Updates proto/news_item.proto formula comment to reflect new weights
- Updates cache-keys.ts documentation
No cache migration needed: the classify cache stores {level, category},
not scores. Scores are computed at read time from the stored level +
the formula, so new digest requests immediately produce new scores.
Gates remain OFF. After 48h of v3 data, re-run:
node scripts/shadow-score-report.mjs
node scripts/shadow-score-rank.mjs sample 25
🤖 Generated with Claude Opus 4.6 via Claude Code + Compound Engineering v2.49.0
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* chore: regenerate proto OpenAPI docs for weight rebalance
* fix(scoring): bump SHADOW_SCORE_LOG_KEY export to v3
The exported constant in cache-keys.ts was left at v2 while the relay's
local constant was bumped to v3. Anyone importing the export (or grep-
discovering it) would get a stale key. Architecture review flagged this.
* fix(scoring): update test + stale comments for shadow-log v3
Review found the regression test still asserted v2 key, causing CI
failure. Also fixed stale v1/v2 references in report script header,
default-key comment, report title render, and shouldNotify docstring.
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
376 lines
16 KiB
YAML
376 lines
16 KiB
YAML
openapi: 3.1.0
|
||
info:
|
||
title: NewsService API
|
||
version: 1.0.0
|
||
paths:
|
||
/api/news/v1/summarize-article:
|
||
post:
|
||
tags:
|
||
- NewsService
|
||
summary: SummarizeArticle
|
||
description: SummarizeArticle generates an LLM summary with provider selection and fallback support.
|
||
operationId: SummarizeArticle
|
||
requestBody:
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/SummarizeArticleRequest'
|
||
required: true
|
||
responses:
|
||
"200":
|
||
description: Successful response
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/SummarizeArticleResponse'
|
||
"400":
|
||
description: Validation error
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/ValidationError'
|
||
default:
|
||
description: Error response
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/Error'
|
||
/api/news/v1/summarize-article-cache:
|
||
get:
|
||
tags:
|
||
- NewsService
|
||
summary: GetSummarizeArticleCache
|
||
description: GetSummarizeArticleCache looks up a cached summary by deterministic key (CDN-cacheable GET).
|
||
operationId: GetSummarizeArticleCache
|
||
parameters:
|
||
- name: cache_key
|
||
in: query
|
||
description: Deterministic cache key computed by buildSummaryCacheKey().
|
||
required: false
|
||
schema:
|
||
type: string
|
||
responses:
|
||
"200":
|
||
description: Successful response
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/SummarizeArticleResponse'
|
||
"400":
|
||
description: Validation error
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/ValidationError'
|
||
default:
|
||
description: Error response
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/Error'
|
||
/api/news/v1/list-feed-digest:
|
||
get:
|
||
tags:
|
||
- NewsService
|
||
summary: ListFeedDigest
|
||
description: ListFeedDigest returns a pre-aggregated digest of all RSS feeds for a site variant.
|
||
operationId: ListFeedDigest
|
||
parameters:
|
||
- name: variant
|
||
in: query
|
||
description: 'Site variant: full, tech, finance, happy'
|
||
required: false
|
||
schema:
|
||
type: string
|
||
- name: lang
|
||
in: query
|
||
description: ISO 639-1 language code (en, fr, ar, etc.)
|
||
required: false
|
||
schema:
|
||
type: string
|
||
responses:
|
||
"200":
|
||
description: Successful response
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/ListFeedDigestResponse'
|
||
"400":
|
||
description: Validation error
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/ValidationError'
|
||
default:
|
||
description: Error response
|
||
content:
|
||
application/json:
|
||
schema:
|
||
$ref: '#/components/schemas/Error'
|
||
components:
|
||
schemas:
|
||
Error:
|
||
type: object
|
||
properties:
|
||
message:
|
||
type: string
|
||
description: Error message (e.g., 'user not found', 'database connection failed')
|
||
description: Error is returned when a handler encounters an error. It contains a simple error message that the developer can customize.
|
||
FieldViolation:
|
||
type: object
|
||
properties:
|
||
field:
|
||
type: string
|
||
description: The field path that failed validation (e.g., 'user.email' for nested fields). For header validation, this will be the header name (e.g., 'X-API-Key')
|
||
description:
|
||
type: string
|
||
description: Human-readable description of the validation violation (e.g., 'must be a valid email address', 'required field missing')
|
||
required:
|
||
- field
|
||
- description
|
||
description: FieldViolation describes a single validation error for a specific field.
|
||
ValidationError:
|
||
type: object
|
||
properties:
|
||
violations:
|
||
type: array
|
||
items:
|
||
$ref: '#/components/schemas/FieldViolation'
|
||
description: List of validation violations
|
||
required:
|
||
- violations
|
||
description: ValidationError is returned when request validation fails. It contains a list of field violations describing what went wrong.
|
||
SummarizeArticleRequest:
|
||
type: object
|
||
properties:
|
||
provider:
|
||
type: string
|
||
minLength: 1
|
||
description: 'LLM provider: "ollama", "groq", "openrouter"'
|
||
headlines:
|
||
type: array
|
||
items:
|
||
type: string
|
||
minItems: 1
|
||
description: Headlines to summarize (max 8 used).
|
||
minItems: 1
|
||
mode:
|
||
type: string
|
||
description: 'Summarization mode: "brief", "analysis", "translate", "" (default).'
|
||
geoContext:
|
||
type: string
|
||
description: Geographic signal context to include in the prompt.
|
||
variant:
|
||
type: string
|
||
description: 'Variant: "full", "tech", or target language for translate mode.'
|
||
lang:
|
||
type: string
|
||
description: Output language code, default "en".
|
||
systemAppend:
|
||
type: string
|
||
description: Optional system prompt append for analytical framework instructions.
|
||
required:
|
||
- provider
|
||
description: SummarizeArticleRequest specifies parameters for LLM article summarization.
|
||
SummarizeArticleResponse:
|
||
type: object
|
||
properties:
|
||
summary:
|
||
type: string
|
||
description: The generated summary text.
|
||
model:
|
||
type: string
|
||
description: Model identifier used for generation.
|
||
provider:
|
||
type: string
|
||
description: Provider that produced the result (or "cache").
|
||
tokens:
|
||
type: integer
|
||
format: int32
|
||
description: Token count from the LLM response.
|
||
fallback:
|
||
type: boolean
|
||
description: Whether the client should try the next provider in the fallback chain.
|
||
error:
|
||
type: string
|
||
description: Error message if the request failed.
|
||
errorType:
|
||
type: string
|
||
description: Error type/name (e.g. "TypeError").
|
||
status:
|
||
type: string
|
||
enum:
|
||
- SUMMARIZE_STATUS_UNSPECIFIED
|
||
- SUMMARIZE_STATUS_SUCCESS
|
||
- SUMMARIZE_STATUS_CACHED
|
||
- SUMMARIZE_STATUS_SKIPPED
|
||
- SUMMARIZE_STATUS_ERROR
|
||
description: SummarizeStatus indicates the outcome of a summarization request.
|
||
statusDetail:
|
||
type: string
|
||
description: Human-readable detail for non-success statuses (skip reason, etc.).
|
||
description: SummarizeArticleResponse contains the LLM summarization result.
|
||
GetSummarizeArticleCacheRequest:
|
||
type: object
|
||
properties:
|
||
cacheKey:
|
||
type: string
|
||
description: Deterministic cache key computed by buildSummaryCacheKey().
|
||
description: GetSummarizeArticleCacheRequest looks up a pre-computed summary by cache key.
|
||
ListFeedDigestRequest:
|
||
type: object
|
||
properties:
|
||
variant:
|
||
type: string
|
||
description: 'Site variant: full, tech, finance, happy'
|
||
lang:
|
||
type: string
|
||
description: ISO 639-1 language code (en, fr, ar, etc.)
|
||
ListFeedDigestResponse:
|
||
type: object
|
||
properties:
|
||
categories:
|
||
type: object
|
||
additionalProperties:
|
||
$ref: '#/components/schemas/CategoryBucket'
|
||
description: Per-category buckets — keys match category names from feed config
|
||
feedStatuses:
|
||
type: object
|
||
additionalProperties:
|
||
type: string
|
||
description: |-
|
||
Per-feed status — only non-ok states emitted; absent key implies ok.
|
||
Values: empty (feed returned 0 items), timeout (timed out during fetch).
|
||
generatedAt:
|
||
type: string
|
||
description: ISO 8601 timestamp of when this digest was generated
|
||
CategoriesEntry:
|
||
type: object
|
||
properties:
|
||
key:
|
||
type: string
|
||
value:
|
||
$ref: '#/components/schemas/CategoryBucket'
|
||
FeedStatusesEntry:
|
||
type: object
|
||
properties:
|
||
key:
|
||
type: string
|
||
value:
|
||
type: string
|
||
CategoryBucket:
|
||
type: object
|
||
properties:
|
||
items:
|
||
type: array
|
||
items:
|
||
$ref: '#/components/schemas/NewsItem'
|
||
NewsItem:
|
||
type: object
|
||
properties:
|
||
source:
|
||
type: string
|
||
minLength: 1
|
||
description: Source feed name.
|
||
title:
|
||
type: string
|
||
minLength: 1
|
||
description: Article headline.
|
||
link:
|
||
type: string
|
||
description: Article URL.
|
||
publishedAt:
|
||
type: integer
|
||
format: int64
|
||
description: 'Publication time, as Unix epoch milliseconds.. Warning: Values > 2^53 may lose precision in JavaScript'
|
||
isAlert:
|
||
type: boolean
|
||
description: Whether this article triggered an alert condition.
|
||
threat:
|
||
$ref: '#/components/schemas/ThreatClassification'
|
||
location:
|
||
$ref: '#/components/schemas/GeoCoordinates'
|
||
locationName:
|
||
type: string
|
||
description: Human-readable location name.
|
||
importanceScore:
|
||
type: integer
|
||
format: int32
|
||
description: 'Composite importance score (0-100): severity × 55% + source tier × 20% + corroboration × 15% + recency × 10%.'
|
||
corroborationCount:
|
||
type: integer
|
||
format: int32
|
||
description: Number of distinct sources that reported the same story in this digest cycle.
|
||
storyMeta:
|
||
$ref: '#/components/schemas/StoryMeta'
|
||
required:
|
||
- source
|
||
- title
|
||
description: NewsItem represents a single news article from RSS feed aggregation.
|
||
ThreatClassification:
|
||
type: object
|
||
properties:
|
||
level:
|
||
type: string
|
||
enum:
|
||
- THREAT_LEVEL_UNSPECIFIED
|
||
- THREAT_LEVEL_LOW
|
||
- THREAT_LEVEL_MEDIUM
|
||
- THREAT_LEVEL_HIGH
|
||
- THREAT_LEVEL_CRITICAL
|
||
description: ThreatLevel represents the assessed threat level of a news event.
|
||
category:
|
||
type: string
|
||
description: Event category.
|
||
confidence:
|
||
type: number
|
||
maximum: 1
|
||
minimum: 0
|
||
format: double
|
||
description: Confidence score (0.0 to 1.0).
|
||
source:
|
||
type: string
|
||
description: Classification source — "keyword", "ml", or "llm".
|
||
description: ThreatClassification represents an AI-assessed threat level for a news item.
|
||
GeoCoordinates:
|
||
type: object
|
||
properties:
|
||
latitude:
|
||
type: number
|
||
maximum: 90
|
||
minimum: -90
|
||
format: double
|
||
description: Latitude in decimal degrees (-90 to 90).
|
||
longitude:
|
||
type: number
|
||
maximum: 180
|
||
minimum: -180
|
||
format: double
|
||
description: Longitude in decimal degrees (-180 to 180).
|
||
description: GeoCoordinates represents a geographic location using WGS84 coordinates.
|
||
StoryMeta:
|
||
type: object
|
||
properties:
|
||
firstSeen:
|
||
type: integer
|
||
format: int64
|
||
description: 'Epoch ms when the story first appeared in any digest cycle.. Warning: Values > 2^53 may lose precision in JavaScript'
|
||
mentionCount:
|
||
type: integer
|
||
format: int32
|
||
description: Total number of digest cycles in which this story appeared.
|
||
sourceCount:
|
||
type: integer
|
||
format: int32
|
||
description: Number of unique sources that reported this story (cached from Redis Set).
|
||
phase:
|
||
type: string
|
||
enum:
|
||
- STORY_PHASE_UNSPECIFIED
|
||
- STORY_PHASE_BREAKING
|
||
- STORY_PHASE_DEVELOPING
|
||
- STORY_PHASE_SUSTAINED
|
||
- STORY_PHASE_FADING
|
||
description: StoryPhase represents the lifecycle stage of a tracked news story.
|
||
description: StoryMeta carries cross-cycle persistence data attached to each news item.
|