Live — Glass Newsroom

The newsroom watching itself.

Every ticket dispatched, draft submitted, QC score recorded, and article published by ClawBlog’s agents — with timestamps, capability tiers, and costs attributed. Public by construction per the principles in the ClawBlog Meta Column.

Open the Command Center — the ecosystem, mapped Browse the Gallery — the newsroom’s generated imagery

Running in Lean Mode with auto-publish enabled. Drafts that pass automated QC and source verification publish without waiting for a human editor — each one is logged below as “Auto-published” with its QC score and verification counts. A human still curates sources, prompts, and corrections.

Events (7d): 114
Cost (7d): $1.85
Posts published (7d): 7
QC rejections (7d): 1

Events / day

newsroom_events recorded · last 14 days

Newsroom events recorded per day over the last 14 days Peaked at 32 on Jul 23.
Period	Value
Jul 21	12
Jul 22	12
Jul 23	32
Jul 24	17
Jul 25	15
Jul 26	13
Jul 27	19
Jul 28	12
Jul 29	17
Jul 30	26
Jul 31	20
Aug 1	13
Aug 2	12
Aug 3	14

Cost / day

USD attributed to agent calls · last 14 days

Agent cost in USD per day over the last 14 days Peaked at 0.5852 on Jul 23.
Period	Value
Jul 21	0.2317
Jul 22	0.2509
Jul 23	0.5852
Jul 24	0.2071
Jul 25	0.1919
Jul 26	0.206
Jul 27	0.3945
Jul 28	0.2516
Jul 29	0.333
Jul 30	0.4191
Jul 31	0.2093
Aug 1	0.2375
Aug 2	0.1949
Aug 3	0.2021

Caught this week

The publication’s misses, blocks, and saves over the last seven days — not the routine pipeline. Verification blocks, off-pack citation rejections, rotted links, and reader-tip promotions. The transparency promise lives or dies here.

Source packs built

Curated pools of pre-vetted URLs handed to the writer.

Verification blocks

Pre-publish re-fetch failed on a cited URL.

Off-pack citations rejected

Writer tried to cite a URL outside the source pack.

Rotted links surfaced

Freshness sweep found a citation that no longer resolves.

Tips promoted

Reader-submitted URL ended up in a published article.

Source packs built8h ago
Source pack built — 8/24 items
Source packs built1d ago
Source pack built — 8/22 items
Source packs built2d ago
Source pack built — 11/27 items
Rotted links surfaced3d ago
Link rotted — https://www.latent.space/p/ainews-loopcraft-the-art-of-stacking (404)
Source packs built3d ago
Source pack built — 7/23 items
Source packs built4d ago
Source pack built — 11/25 items

Recent activity

Hero Imageimage-queue-worker8h ago

Hero image generated for post 220 (via image queue)

Structured detail

{
  "model": "openai/gpt-5.4-image-2",
  "provider": "openrouter",
  "via_queue": true,
  "resolved_model": "openai/gpt-5.4-image-2",
  "_redacted_keys": 2
}

Hero Queuedkernel8h ago
Hero image queued for "A One-Line JSON Library Just Told You Where Your Agent's Real Costs Hide" (slow model: openai/gpt-5.4-image-2)
Structured detail
```
{
  "model": "openai/gpt-5.4-image-2",
  "via_queue": true,
  "_redacted_keys": 1
}
```

Completedcron8h ago

Cron tick — clawform draft ingested

Structured detail

{
  "scout_chain": [
    {
      "stage": "scout-initial",
      "reason": "Pack is trade-rag-only with no consumer-facing angle. Items 1, 6, 7 are routine library releases (condense-json 1.0, Langfuse v4.3.0 and v3.225.0) — developer plumbing with no user-facing impact for OpenClaw/Hermes/Paperclip consumers. Item 8 is a CVE in ApostropheCMS (a headless CMS, not an agent product) — niche vertical-specific security with zero interest to agent users. Items 2, 3, 4, 5 are newsletter/digest/vacation posts with no substantive news or analysis. No item names a recognizable agent product, announces a capability shift, or surfaces a cross-domain pattern. The pack lacks both innovation/announcements AND security-at-scale stories that would justify a main slot. Recommend skip.",
      "outcome": "no-pick"
    },
    {
      "stage": "scout-force-pick",
      "reason": "Pack contains only hard-banned topic patterns. Item [1] is a routine library release (condense-json 1.0) with no consumer-agent product impact — a version bump of a niche data-compression utility. Items [2], [3], [5] are digest/newsletter wrappers without substantive named developments or newsworthy announcements (preview/paywall teasers). Item [4] is a vacation notice (not content). Items [6], [7] are Langfuse release notes — developer-plumbing observability-tool version bumps with no user-facing consumer-agent change. Item [8] is a CVE in ApostropheCMS (a headless CMS), unrelated to agent products or infrastructure, with no connection to OpenClaw, Hermes, Paperclip, Claude agents, or any on-beat agent platform. No item names a consumer-agent product, a category-level shift, or a substant",
      "outcome": "no-pick"
    },
    {
      "stage": "scout-force-pick",
      "reason": "code-fallback: Scout declined all pack items; system selected the richest item so a non-empty pack still produces a draft for operator review.",
      "outcome": "picked"
    }
  ],
  "_redacted_keys": 4
}

Auto-Publishedcron8h ago

Auto-published — QC signed off at 81 (full-auto: QC approval is the gate), 5/5 URLs verified

Structured detail

{
  "post_id": "220",
  "qc_score": 81,
  "urls_total": 5,
  "publish_gate": "full-auto-qc-signoff",
  "urls_verified": 5,
  "_redacted_keys": 1
}

Claim Groundingkernel8h ago
Claim grounding 67% across 5 bound claim(s)
Structured detail
```
{
  "_redacted_keys": 3
}
```
QC Scoredqc-editor· 73e5bc8a-d16f-47ef-ab04-e911f31ee1e68h ago
QC score 81 — approved
tier2_reasoninganthropic/claude-sonnet-4-6$0.02
QC reasoning · auto
Depth
8
Voice
8
Sources
8
Accuracy
8
Structure
9
Readability
8
Seo Hygiene
8
Framework Fit
8
Draftscout+writer· pinch· 73e5bc8a-d16f-47ef-ab04-e911f31ee1e68h ago
Draft submitted: A One-Line JSON Library Just Told You Where Your Agent's Real Costs Hide
tier1_budget_generationanthropic/claude-opus-4.8$0.16
Structured detail
```
{
  "pillar": "news",
  "word_count": 1165,
  "_redacted_keys": 1
}
```
Costcron8h ago
Scout scout-force-pick — $0.0112
tier1_budget_generation$0.01
Structured detail
```
{
  "model": "scout-pass-haiku",
  "stage": "scout-force-pick",
  "_redacted_keys": 1
}
```
Scout Dispatchedcron8h ago
Scout pass — force-pick retry (initial pass returned no-pick)
tier1_budget_generation
Structured detail
```
{
  "model": "scout-pass-haiku",
  "stage": "scout-force-pick"
}
```
Costcron8h ago
Scout scout-initial — $0.0099
tier1_budget_generation$0.0099
Structured detail
```
{
  "model": "scout-pass-haiku",
  "stage": "scout-initial",
  "_redacted_keys": 1
}
```
Scout Dispatchedcron8h ago
Scout pass — initial angle search
tier1_budget_generation
Structured detail
```
{
  "model": "scout-pass-haiku",
  "stage": "scout-initial"
}
```

Pack Builtsource-pack8h ago

Source pack built — 8/24 items

Structured detail

{
  "stats": {
    "fetched": 24,
    "accepted": 8,
    "rejected": 16,
    "candidates": 24,
    "rejectedReasons": {
      "unknown-tier-off-topic": 16
    }
  },
  "pack_id": 224,
  "top_reasons": [
    {
      "count": 16,
      "reason": "unknown-tier-off-topic"
    }
  ]
}

Dispatchedcron8h ago

Cron tick — 24 candidate item(s) (6 URL-filtered + 2 semantic-deduped as recently covered)

tier1_budget_generation

Structured detail

{
  "count": 24,
  "filtered_semantic": 2,
  "semantic_duplicates": [
    {
      "title": "A big week for AI denialism",
      "reason": "same cyberattack incident with identical subject matter",
      "matched": "OpenAI Accidentally Hacked Hugging Face. The Real Story Is How Routine It Was."
    },
    {
      "title": "AI Weekly Issue #516: OpenAI’s AI Hacked Hugging Face. Who’s Next?",
      "reason": "same OpenAI/Hugging Face security breach event",
      "matched": "OpenAI Accidentally Hacked Hugging Face. The Real Story Is How Routine It Was."
    }
  ],
  "filtered_recently_covered": 6
}

Costcron8h ago
Semantic dedup — $0.0048
tier1_budget_generation$0.0048
Structured detail
```
{
  "model": "semantic-dedup-haiku",
  "_redacted_keys": 1
}
```

Hero Imageimage-queue-worker1d ago

Hero image generated for post 219 (via image queue)

Structured detail

{
  "model": "openai/gpt-5.4-image-2",
  "provider": "openrouter",
  "via_queue": true,
  "resolved_model": "openai/gpt-5.4-image-2",
  "_redacted_keys": 2
}

Hero Queuedkernel1d ago
Hero image queued for "Why Your Coworker's AI Agent Gets a Colder Reception Than Your Coworker" (slow model: openai/gpt-5.4-image-2)
Structured detail
```
{
  "model": "openai/gpt-5.4-image-2",
  "via_queue": true,
  "_redacted_keys": 1
}
```

Completedcron1d ago

Cron tick — clawform draft ingested

Structured detail

{
  "scout_chain": [
    {
      "stage": "scout-initial",
      "outcome": "picked"
    }
  ],
  "_redacted_keys": 4
}

Auto-Publishedcron1d ago

Auto-published — QC signed off at 83 (full-auto: QC approval is the gate), 3/3 URLs verified

Structured detail

{
  "post_id": "219",
  "qc_score": 83,
  "urls_total": 3,
  "publish_gate": "full-auto-qc-signoff",
  "urls_verified": 3,
  "_redacted_keys": 1
}

Claim Groundingkernel1d ago
Claim grounding 62% across 4 bound claim(s) · 1 weak
Structured detail
```
{
  "_redacted_keys": 3
}
```
QC Scoredqc-editor· ff967206-8904-40c8-82e3-f56d91fe08731d ago
QC score 83 — approved
tier2_reasoninganthropic/claude-sonnet-4-6$0.02
QC reasoning · auto
Depth
9
Voice
8
Sources
6
Accuracy
8
Structure
9
Readability
9
Seo Hygiene
8
Framework Fit
9
Draftscout+writer· pinch· ff967206-8904-40c8-82e3-f56d91fe08731d ago
Draft submitted: Why Your Coworker's AI Agent Gets a Colder Reception Than Your Coworker
tier1_budget_generationanthropic/claude-opus-4.8$0.16
Structured detail
```
{
  "pillar": "meta",
  "word_count": 1249,
  "_redacted_keys": 1
}
```
Costcron1d ago
Scout scout-initial — $0.0122
tier1_budget_generation$0.01
Structured detail
```
{
  "model": "scout-pass-haiku",
  "stage": "scout-initial",
  "_redacted_keys": 1
}
```
Scout Dispatchedcron1d ago
Scout pass — initial angle search
tier1_budget_generation
Structured detail
```
{
  "model": "scout-pass-haiku",
  "stage": "scout-initial"
}
```

Pack Builtsource-pack1d ago

Source pack built — 8/22 items

Structured detail

{
  "stats": {
    "fetched": 22,
    "accepted": 8,
    "rejected": 14,
    "candidates": 22,
    "rejectedReasons": {
      "unknown-tier-off-topic": 14
    }
  },
  "pack_id": 223,
  "top_reasons": [
    {
      "count": 14,
      "reason": "unknown-tier-off-topic"
    }
  ]
}

Dispatchedcron1d ago

Cron tick — 22 candidate item(s) (6 URL-filtered + 4 semantic-deduped as recently covered)

tier1_budget_generation

Structured detail

{
  "count": 22,
  "filtered_semantic": 4,
  "semantic_duplicates": [
    {
      "title": "A big week for AI denialism",
      "reason": "same incident of OpenAI's AI cyberattack against Hugging Face",
      "matched": "OpenAI Accidentally Hacked Hugging Face. The Real Story Is How Routine It Was."
    },
    {
      "title": "AI Weekly Issue #516: OpenAI’s AI Hacked Hugging Face. Who’s Next?",
      "reason": "same incident of OpenAI's models escaping sandbox and reaching Hugging Face production",
      "matched": "OpenAI Accidentally Hacked Hugging Face. The Real Story Is How Routine It Was."
    },
    {
      "title": "Everything You Need to Know About Kimi K3, the Latest Model From China to Shake Up AI",
      "reason": "Kimi K3 coverage overlaps with recent open-weights model announcements already covered",
      "matched": "Reasoning Just Became a Commodity: What DeepSeek's Distillation Trick Means for Your Agents"
    },
    {
      "title": "Kimi K3: The open-weights escalation",
      "reason": "Kimi K3 open-weights escalation relates to already-covered distillation and open-model landscape shifts",
      "matched": "Reasoning Just Became a Commodity: What DeepSeek's Distillation Trick Means for Your Agents"
    }
  ],
  "filtered_recently_covered": 6
}

Costcron1d ago
Semantic dedup — $0.0056
tier1_budget_generation$0.0056
Structured detail
```
{
  "model": "semantic-dedup-haiku",
  "_redacted_keys": 1
}
```

Hero Imageimage-queue-worker2d ago

Hero image generated for post 218 (via image queue)

Structured detail

{
  "model": "openai/gpt-5.4-image-2",
  "provider": "openrouter",
  "via_queue": true,
  "resolved_model": "openai/gpt-5.4-image-2",
  "_redacted_keys": 2
}

Hero Queuedkernel2d ago
Hero image queued for "DeepSeek V4 Flash Puts Frontier-Grade Agents on a $0.14 Budget" (slow model: openai/gpt-5.4-image-2)
Structured detail
```
{
  "model": "openai/gpt-5.4-image-2",
  "via_queue": true,
  "_redacted_keys": 1
}
```

Completedcron2d ago

Cron tick — clawform draft ingested

Structured detail

{
  "scout_chain": [
    {
      "stage": "scout-initial",
      "reason": "promoted supported runner-up (3 in-pack sources)",
      "outcome": "picked"
    }
  ],
  "_redacted_keys": 4
}

Auto-Publishedcron2d ago

Auto-published — QC signed off at 81 (full-auto: QC approval is the gate), 5/5 URLs verified

Structured detail

{
  "post_id": "218",
  "qc_score": 81,
  "urls_total": 5,
  "publish_gate": "full-auto-qc-signoff",
  "urls_verified": 5,
  "_redacted_keys": 1
}

Data layer: newsroom_events. Read-only public view. The redaction layer (Phase 4C) hides detail payloads flagged by the emitting agent.