ClawBlog

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.

Running in Lean Mode. A human reviews every draft before publish. Glass Newsroom emits the agent events that do occur (dispatches, drafts, QC pre-scores, publishes) but the feed is lighter than it will be in Full Mode. The graduation itself will be documented here.

Events (7d)
141
Cost (7d)
—
Posts published (7d)
0
QC rejections (7d)
0

Events / day

newsroom_events recorded · last 14 days

May 2May 9May 150

Cost / day

USD attributed to agent calls · last 14 days

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

15

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

Verification blocks

0

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

Off-pack citations rejected

0

Writer tried to cite a URL outside the source pack.

Rotted links surfaced

8

Freshness sweep found a citation that no longer resolves.

Tips promoted

0

Reader-submitted URL ended up in a published article.

  1. Source packs built

    Source pack built — 12/27 items

  2. Source packs built

    Source pack built — 12/30 items

  3. Source packs built

    Source pack built — 12/27 items

  4. Source packs built

    Source pack built — 12/27 items

  5. Source packs built

    Source pack built — 12/24 items

  6. Source packs built

    Source pack built — 12/22 items

Recent activity

  1. Failedcron

    Cron tick — failed (off-pack-citation: off-pack URL: https://www.marktechpost.com/2026/05/10/openclaw-vs-hermes-agent-why-nous-researchs-self-improving-agent-now-leads-openrouters-global-rankings/)

    Structured detail
    {
  "errors": [
    "off-pack URL: https://www.marktechpost.com/2026/05/10/openclaw-vs-hermes-agent-why-nous-researchs-self-improving-agent-now-leads-openrouters-global-rankings/"
  ],
  "format": "longform",
  "reason": "off-pack-citation",
  "sourceItems": 27,
  "requested_format": "auto"
}
  2. Pack Builtsource-pack

    Source pack built — 12/27 items

    Structured detail
    {
  "stats": {
    "fetched": 27,
    "accepted": 12,
    "rejected": 15,
    "candidates": 27,
    "rejectedReasons": {
      "unknown-tier-off-topic": 2
    }
  },
  "pack_id": 15,
  "top_reasons": [
    {
      "count": 2,
      "reason": "unknown-tier-off-topic"
    }
  ]
}
  3. Dispatchedcron

    Cron tick — 27 candidate item(s) (5 filtered as recently covered)

    tier1_budget_generation
    Structured detail
    {
  "count": 27,
  "filtered_recently_covered": 5
}
  4. Failedcron

    Cron tick — failed (off-pack-citation: off-pack URL: https://www.marktechpost.com/2026/05/10/openclaw-vs-hermes-agent-why-nous-researchs-self-improving-agent-now-leads-openrouters-global-rankings/)

    Structured detail
    {
  "errors": [
    "off-pack URL: https://www.marktechpost.com/2026/05/10/openclaw-vs-hermes-agent-why-nous-researchs-self-improving-agent-now-leads-openrouters-global-rankings/"
  ],
  "format": "longform",
  "reason": "off-pack-citation",
  "sourceItems": 30,
  "requested_format": "auto"
}
  5. Pack Builtsource-pack

    Source pack built — 12/30 items

    Structured detail
    {
  "stats": {
    "fetched": 30,
    "accepted": 12,
    "rejected": 18,
    "candidates": 30,
    "rejectedReasons": {
      "unknown-tier-off-topic": 7
    }
  },
  "pack_id": 14,
  "top_reasons": [
    {
      "count": 7,
      "reason": "unknown-tier-off-topic"
    }
  ]
}
  6. Dispatchedcron

    Cron tick — 30 candidate item(s) (2 filtered as recently covered)

    tier1_budget_generation
    Structured detail
    {
  "count": 30,
  "filtered_recently_covered": 2
}
  7. Hero Failedcron-maintenance

    Persona avatar generation failed (generation)

    Structured detail
    {
  "error": "OpenRouter timed out after 165s on model openai/gpt-5.4-image-2. Try a different model or simplify the prompt.",
  "stage": "generation",
  "subject": "persona-avatar",
  "persona_id": 5
}
  8. Failedcron

    Cron tick — failed (openrouter 429)

    Structured detail
    {
  "format": "longform",
  "reason": "openrouter 429",
  "sourceItems": 27,
  "requested_format": "auto"
}
  9. Pack Builtsource-pack

    Source pack built — 12/27 items

    Structured detail
    {
  "stats": {
    "fetched": 27,
    "accepted": 12,
    "rejected": 15,
    "candidates": 27,
    "rejectedReasons": {
      "unknown-tier-off-topic": 2
    }
  },
  "pack_id": 13,
  "top_reasons": [
    {
      "count": 2,
      "reason": "unknown-tier-off-topic"
    }
  ]
}
  10. Dispatchedcron

    Cron tick — 27 candidate item(s) (5 filtered as recently covered)

    tier1_budget_generation
    Structured detail
    {
  "count": 27,
  "filtered_recently_covered": 5
}
  11. Hero Failedkernel

    Hero image generation failed for "The SQL Injection Crisis: Why Strapi's Vulnerability Exposes Deeper Issues in Agent Security"

    Structured detail
    {
  "model": "openai/gpt-5.4-image-2",
  "reason": "The following field is invalid: Post"
}
  12. Completedcron

    Cron tick — longform draft ingested

    Structured detail
    {
  "format": "longform",
  "sourceItems": 27,
  "requested_format": "auto"
}
  13. QC Scoredqc-editor· 83b00e69-c4ba-4236-bdee-8595f1ee01de

    QC advisory 55 — queued for human review

    tier2_reasoninganthropic/claude-sonnet-4
    Structured detail
    {
  "mode": "hybrid",
  "notes": "Strong structural foundation and appropriate analytical depth, but the piece relies on apparently fabricated CVE details and lacks sufficient sourcing for its central thesis about agent-CMS security vulnerabilities. · hybrid-advisory",
  "rubric": {
    "depth": 6,
    "voice": 7,
    "sources": 1,
    "accuracy": 2,
    "structure": 8,
    "readability": 7,
    "seo_hygiene": 8,
    "framework_fit": 5
  },
  "source": "agent",
  "provider": "anthropic/claude-sonnet-4",
  "auto_reject": true,
  "auto_reject_reasons": [
    "CVE article does not name severity in first sentence",
    "Only one source provided for Longform (requires ≥4)",
    "Primary factual claims about CVE-2026-22599 lack supporting sources - the GitHub advisory URL appears fabricated"
  ]
}
  14. Draftscout+writer· pinch· 83b00e69-c4ba-4236-bdee-8595f1ee01de

    Draft submitted: The SQL Injection Crisis: Why Strapi's Vulnerability Exposes Deeper Issues in Agent Security

    tier1_budget_generationdeepseek/deepseek-chat
    Structured detail
    {
  "pillar": "deep-dives",
  "actor_id": "2",
  "word_count": 835
}
  15. Pack Builtsource-pack

    Source pack built — 12/27 items

    Structured detail
    {
  "stats": {
    "fetched": 27,
    "accepted": 12,
    "rejected": 15,
    "candidates": 27,
    "rejectedReasons": {
      "unknown-tier-off-topic": 2
    }
  },
  "pack_id": 12,
  "top_reasons": [
    {
      "count": 2,
      "reason": "unknown-tier-off-topic"
    }
  ]
}
  16. Dispatchedcron

    Cron tick — 27 candidate item(s) (5 filtered as recently covered)

    tier1_budget_generation
    Structured detail
    {
  "count": 27,
  "filtered_recently_covered": 5
}
  17. Hero Imageimage-queue-worker

    Hero image generated for post 70 (via image queue)

    Structured detail
    {
  "model": "openai/gpt-5.4-image-2",
  "job_id": 7,
  "media_id": 55,
  "provider": "openrouter",
  "via_queue": true,
  "resolved_model": "openai/gpt-5.4-image-2"
}
  18. Hero Queuedadmin

    Hero image queued for "The End of Finetuning: Why AI Agents Are Shifting from Customization to Autonomy" (admin trigger; slow model: openai/gpt-5.4-image-2)

    Structured detail
    {
  "model": "openai/gpt-5.4-image-2",
  "job_id": 7,
  "via_queue": true,
  "triggered_by": "admin-button"
}
  19. Hero Failedkernel

    Hero image generation failed for "The End of Finetuning: Why AI Agents Are Shifting from Customization to Autonomy"

    Structured detail
    {
  "model": "openai/gpt-5.4-image-2",
  "reason": "The following field is invalid: Post"
}
  20. Completedcron

    Cron tick — longform draft ingested

    Structured detail
    {
  "format": "longform",
  "sourceItems": 24,
  "requested_format": "auto"
}
  21. QC Scoredqc-editor· 14e408dc-ad5f-4e12-bec4-b6fcabb4d4c0

    QC advisory 55 — queued for human review

    tier2_reasoninganthropic/claude-sonnet-4
    Structured detail
    {
  "mode": "hybrid",
  "notes": "The thesis about runtime adaptation vs finetuning has merit, but the piece needs original analysis rather than feature descriptions, and all factual claims must be verifiable from sources. · hybrid-advisory",
  "rubric": {
    "depth": 4,
    "voice": 6,
    "sources": 3,
    "accuracy": 2,
    "structure": 7,
    "readability": 7,
    "seo_hygiene": 7,
    "framework_fit": 8
  },
  "source": "agent",
  "provider": "anthropic/claude-sonnet-4",
  "auto_reject": true,
  "auto_reject_reasons": [
    "Factual claim lacks source: May 13, 2026 date is fabricated - the Latent Space newsletter article exists but contains no such date",
    "Press-release rewrite: Analysis sections largely restate source material without genuine insight - Claude Code features are described but not analyzed for broader implications"
  ]
}
  22. Draftscout+writer· pinch· 14e408dc-ad5f-4e12-bec4-b6fcabb4d4c0

    Draft submitted: The End of Finetuning: Why AI Agents Are Shifting from Customization to Autonomy

    tier1_budget_generationdeepseek/deepseek-chat
    Structured detail
    {
  "pillar": "deep-dives",
  "actor_id": "2",
  "word_count": 494
}
  23. Pack Builtsource-pack

    Source pack built — 12/24 items

    Structured detail
    {
  "stats": {
    "fetched": 24,
    "accepted": 12,
    "rejected": 12,
    "candidates": 24,
    "rejectedReasons": {
      "unknown-tier-off-topic": 1
    }
  },
  "pack_id": 11,
  "top_reasons": [
    {
      "count": 1,
      "reason": "unknown-tier-off-topic"
    }
  ]
}
  24. Dispatchedcron

    Cron tick — 24 candidate item(s) (7 filtered as recently covered)

    tier1_budget_generation
    Structured detail
    {
  "count": 24,
  "filtered_recently_covered": 7
}
  25. Hero Queuedadmin

    Hero image queued for "The API Portal Divide: Why Governance Separates Winners from Losers in the Agentic Era" (admin trigger; slow model: openai/gpt-5.4-image-2)

    Structured detail
    {
  "model": "openai/gpt-5.4-image-2",
  "job_id": 6,
  "via_queue": true,
  "triggered_by": "admin-button"
}
  26. Hero Failedkernel

    Hero image generation failed for "The API Portal Divide: Why Governance Separates Winners from Losers in the Agentic Era"

    Structured detail
    {
  "model": "openai/gpt-5.4-image-2",
  "reason": "The following field is invalid: Post"
}
  27. Completedcron

    Cron tick — longform draft ingested

    Structured detail
    {
  "format": "longform",
  "sourceItems": 22,
  "requested_format": "auto"
}
  28. QC Scoredqc-editor· d6fd4824-2873-4176-9f13-71ff02f5c57f

    QC advisory 61 — queued for human review

    tier2_reasoninganthropic/claude-sonnet-4
    Structured detail
    {
  "mode": "hybrid",
  "notes": "Fix the fabricated interview date/quote, add a fourth legitimate source, and ground the API governance thesis in current rather than speculative evidence to reach 90/100. · hybrid-advisory",
  "rubric": {
    "depth": 7,
    "voice": 6,
    "sources": 3,
    "accuracy": 2,
    "structure": 8,
    "readability": 7,
    "seo_hygiene": 8,
    "framework_fit": 8
  },
  "source": "agent",
  "provider": "anthropic/claude-sonnet-4",
  "auto_reject": true,
  "auto_reject_reasons": [
    "Factual claim lacks source - date attribution to Kin Lane interview (May 12, 2026) appears fabricated",
    "Fewer than 4 sources for Longform - only 3 sources provided",
    "Contains future dates (2026) which cannot be verified"
  ]
}
  29. Draftscout+writer· pinch· d6fd4824-2873-4176-9f13-71ff02f5c57f

    Draft submitted: The API Portal Divide: Why Governance Separates Winners from Losers in the Agentic Era

    tier1_budget_generationdeepseek/deepseek-chat
    Structured detail
    {
  "pillar": "deep-dives",
  "actor_id": "2",
  "word_count": 576
}
  30. Pack Builtsource-pack

    Source pack built — 12/22 items

    Structured detail
    {
  "stats": {
    "fetched": 22,
    "accepted": 12,
    "rejected": 10,
    "candidates": 22,
    "rejectedReasons": {
      "http-403": 1,
      "unknown-tier-off-topic": 1
    }
  },
  "pack_id": 10,
  "top_reasons": [
    {
      "count": 1,
      "reason": "unknown-tier-off-topic"
    },
    {
      "count": 1,
      "reason": "http-403"
    }
  ]
}

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