Files
llm-wiki/log.md
T
2026-07-18 10:09:57 +09:00

442 lines
48 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# Wiki Log
> Chronological record of all wiki actions. Append-only.
> Format: `## [YYYY-MM-DD] action | subject`
> Actions: ingest, update, query, lint, create, archive, delete
> Previous 2026 entries rotated to `log-2026.md` on 2026-07-03 because the log exceeded 500 entries.
## [2026-07-03] lint | Log rotated
- Rotated previous `log.md` to `log-2026.md` after it reached 767 lines.
- Started a fresh `log.md` before appending the current scheduled Discord link-ingest run.
## [2026-07-03] ingest | Discord link ingest blocked by disk-full share import
- Oriented with `SCHEMA.md`, `index.md`, recent log, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` reported 164,616 local messages, share enabled at `/var/lib/hermes/.local/share/discrawl/share`, and `needs_update=true`.
- The scheduled read-only `discrawl --json sql` path allowed the configured git-share update. This time the git pull/fetch phase progressed, but import failed while clearing/reloading messages with `database or disk is full (13)`.
- Disk evidence: `/var/lib/hermes` is 100% used with about 15M available; `/var/lib/hermes/.local/share/discrawl` is about 3.4G; the git share contains about 2.49 GiB of loose objects and 140 MiB packed objects.
- Fallback read-only Python/SQLite query against the local database found 0 URL messages newer than `2026-07-02T14:22:19.033000000Z` in #chat/#tw. Local max URL timestamps remain #chat `2026-07-02T12:46:45.961000000Z` and #tw `2026-07-02T14:22:19.033000000Z`.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. State updated: `.automation/discord-link-ingest/state.md`.
- Rubric unchanged. Newer Discord share messages cannot be curated until the Hermes volume has enough space for discrawl share import, or the discrawl share storage is compacted/relocated.
## [2026-07-03] ingest | 山本美香
- User-directed ingest from Discord thread prompt: `[toymaker] 山本美香 ingest`.
- Saved raw sources:
- `raw/articles/yamamoto-mika-memorial-foundation-profile-2026.md`
- `raw/articles/yamamoto-mika-memorial-foundation-about-2026.md`
- `raw/articles/cpj-mika-yamamoto-2026.md`
- `raw/articles/suginami-yamamoto-mika-2026.md`
- Created `entities/yamamoto-mika.md`.
- Updated `index.md` total from 52 to 53 pages.
## [2026-07-03] ingest | Discord link ingest blocked by no-space git-share fetch
- Oriented with `SCHEMA.md`, `index.md`, recent log, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` reported 164,616 local messages, share enabled at `/var/lib/hermes/.local/share/discrawl/share`, and `needs_update=true`.
- The scheduled read-only `discrawl --json sql` path allowed the configured git-share update, but `git fetch --prune origin` failed before import with `unable to write loose object file: No space left on device` and `unpack-objects failed`.
- Disk evidence: `/var/lib/hermes` is 100% used with about 12M available; `/var/lib/hermes/.local/share/discrawl` is about 3.4G.
- Fallback read-only Python/SQLite query against the local database found 0 URL messages newer than `2026-07-02T14:22:19.033000000Z` in #chat/#tw. Local max URL timestamps remain #chat `2026-07-02T12:46:45.961000000Z` and #tw `2026-07-02T14:22:19.033000000Z`.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. State updated: `.automation/discord-link-ingest/state.md`.
- Rubric unchanged. Newer Discord share messages cannot be curated until the Hermes volume has enough space for discrawl share fetch/import, or the discrawl share storage is compacted/relocated.
## [2026-07-04] ingest | Discord link ingest blocked by disk-full SQLite I/O error
- Oriented with `SCHEMA.md`, `index.md`, recent log, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` failed with `ping sqlite readonly: disk I/O error (4874)`.
- Disk evidence: `/var/lib/hermes` is 100% used with 0 available; `/tmp` has 1.7G available; `/var/lib/hermes/.local/share/discrawl` is about 3.4G.
- Fallback immutable read-only SQLite query against the local database succeeded and found 0 URL messages newer than `2026-07-02T14:22:19.033000000Z` in #chat/#tw. Local max URL timestamps remain #chat `2026-07-02T12:46:45.961000000Z` and #tw `2026-07-02T14:22:19.033000000Z`.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. State updated: `.automation/discord-link-ingest/state.md`.
- Rubric unchanged. Newer Discord share messages cannot be curated until the Hermes volume has free space for discrawl status/fetch/import, or the discrawl share storage is compacted/relocated.
## [2026-07-04] ingest | Discord link ingest still blocked by disk-full SQLite I/O error
- Oriented with `SCHEMA.md`, `index.md`, recent log, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` again failed with `ping sqlite readonly: disk I/O error (4874)`.
- Disk evidence: `/var/lib/hermes` is 100% used with 0 available; `/tmp` has 1.7G available; `/var/lib/hermes/.local/share/discrawl` is about 3.4G; its git share is about 2.7G.
- Fallback immutable read-only SQLite query against the local database succeeded and found 0 URL messages newer than `2026-07-02T14:22:19.033000000Z` in #chat/#tw. Local max URL timestamps remain #chat `2026-07-02T12:46:45.961000000Z` and #tw `2026-07-02T14:22:19.033000000Z`.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. State updated: `.automation/discord-link-ingest/state.md`.
- Rubric unchanged. Newer Discord share messages cannot be curated until the Hermes volume has free space for discrawl status/fetch/import, or the discrawl share storage is compacted/relocated.
## [2026-07-04] ingest | Discord link ingest still blocked by disk-full SQLite I/O error
- Oriented with `SCHEMA.md`, `index.md`, recent log, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` again failed with `ping sqlite readonly: disk I/O error (4874)`.
- Disk evidence: `/var/lib/hermes` is 100% used with 0 available; `/tmp` has 1.7G available; `/var/lib/hermes/.local/share/discrawl` is about 3.4G.
- Fallback immutable read-only SQLite query against the local database succeeded and found 0 URL messages newer than `2026-07-02T14:22:19.033000000Z` in #chat/#tw. Local max URL timestamps remain #chat `2026-07-02T12:46:45.961000000Z` and #tw `2026-07-02T14:22:19.033000000Z`.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. State updated: `.automation/discord-link-ingest/state.md`.
- Rubric unchanged. Newer Discord share messages cannot be curated until the Hermes volume has free space for discrawl status/fetch/import, or the discrawl share storage is compacted/relocated.
## [2026-07-16] ingest | Discord-discovered agent/security sources
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded; the configured git share auto-update/import ran and imported 167,035 messages from the share.
- Scanned 402 URL-containing #chat/#tw messages newer than `2026-07-02T14:22:19.033000000Z`; extracted 2,230 normalized URLs.
- Raw sources saved: `raw/articles/asyncapi-miasma-supply-chain-worm-2026.md`, `raw/articles/simon-grok-build-source-privacy-2026.md`, `raw/articles/prompt-injection-as-role-confusion-2026.md`, `raw/articles/github-actions-workflow-execution-protections-2026.md`, `raw/articles/chrome-web-identity-io26-2026.md`.
- Wiki page created: `concepts/prompt-injection-role-confusion.md`.
- Wiki pages updated: `concepts/open-source-package-supply-chain-attacks.md`, `concepts/ai-agent-telemetry-privacy.md`, `concepts/ai-agent-command-safety.md`, `concepts/ci-cd-runtime-security.md`, `index.md`.
- Automation files updated: `.automation/discord-link-ingest/state.md`, `.automation/discord-link-ingest/interest-profile.md`.
- Strict skips/link-only: 2,224 URLs, mostly X/Twitter, `t.co`, media, market/sports/geopolitics, and low-context direct links. One attempted arXiv PDF extraction failed because the configured `web_extract` backend is search-only.
## [2026-07-16] ingest | Discord link ingest X-only digest skipped
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. The configured git share auto-update/import path also ran during the SQL read, importing 175,978 share rows; archive now reports 167,078 messages across 261 channels.
- Scanned 4 URL-containing #tw messages newer than `2026-07-16T06:58:33.736000000Z`; extracted 48 URLs / 35 normalized unique URLs.
- All 35 URLs were `x.com` links from a digest. Candidate themes included Codex usage-limit design, Codex parallel scheduling, Sakana/NVIDIA physical AI, Cursor Windows zero-day, and ModelScope/Qwen Code local-LLM discussion.
- Raw sources saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T07:55:15.382000000Z`.
- Rubric unchanged: strict threshold kept X-only discovery links as link-only until a durable primary source/article/repo/docs page appears.
## [2026-07-16] ingest | Discord link ingest X-only chat links skipped
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` then ran the configured git share auto-update/import path, importing 173,537 share rows.
- Scanned 2 URL-containing #chat messages newer than `2026-07-16T07:55:15.382000000Z`; extracted 2 normalized unique URLs.
- Both URLs were `x.com/i/status/...` links. One had context “経営陣がデリバリーまでのフローを分かってると心強い”, but neither exposed a durable primary source/article/repo/docs page.
- Raw sources saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T09:22:19.795000000Z`.
- Rubric unchanged: strict threshold kept X-only discovery links as link-only.
## [2026-07-16] ingest | Discord link ingest X-only digest skipped
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,604 share rows.
- Scanned 4 URL-containing #tw messages newer than `2026-07-16T09:22:19.795000000Z`; extracted 68 URLs / 50 normalized unique URLs.
- All 50 normalized URLs were `x.com` discovery links. Candidate themes included Japan AI infrastructure investment, NVIDIA/Noetra/physical AI, Thinking Machines Lab Inkling, Grok Build, AI tool rate-limit operations, formal/spec-driven development with Lean/Alloy/TLA+/Z3, Manus AI RCE, IoT/AWS security, and cashless-payment reliability incidents.
- Raw sources saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T10:55:44.201000000Z`.
- Rubric unchanged: strict threshold kept X-only digest links as link-only until a durable primary source/article/repo/docs page appears.
## [2026-07-16] ingest | Discord link ingest no new URLs
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,630 share rows.
- Scanned 1 total #chat/#tw message newer than `2026-07-16T10:55:44.201000000Z`; found 0 URL-containing messages.
- Raw sources saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T12:01:50.645000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-16] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,646 share rows.
- Scanned 0 total #chat/#tw messages newer than `2026-07-16T12:01:50.645000000Z`; found 0 URL-containing messages.
- Raw sources saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-16T12:01:50.645000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-16] ingest | 1Password Credential Broker from Discord links
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,679 share rows.
- Scanned 9 total new #chat/#tw messages newer than `2026-07-16T12:01:50.645000000Z`; 6 messages contained URLs.
- Extracted 44 URL mentions / 38 normalized unique URLs. Most were X/Twitter discovery links from the #tw digest or a Spotify/media link and were kept link-only/skipped.
- Source saved: `raw/articles/1password-credential-broker-2026.md` from `https://1password.com/blog/introducing-1password-credential-broker` — score 4, primary source for runtime credential brokering, GitHub Actions workload identity, audit attribution, and future AI-agent short-lived token flow.
- Wiki pages updated: `concepts/ai-agent-identity-security.md`, `concepts/ci-cd-runtime-security.md`.
- Wiki pages created: 0. Raw sources saved: 1. Link-only/skipped: 37. Extraction blockers: web_search backend unavailable; t.co shortlinks stayed link-only rather than fabricating resolved sources.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T14:45:46.580000000Z`.
- Rubric unchanged: the kept source confirms the existing agent identity / credential leakage / CI governance interest.
## [2026-07-16] ingest | Discord link ingest no new URLs
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,691 share rows.
- Scanned 2 total #chat/#tw messages newer than `2026-07-16T14:45:46.580000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T14:53:01.389000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-16] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 689 changed share rows.
- Scanned 0 total #chat/#tw messages newer than `2026-07-16T14:53:01.389000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-16T14:53:01.389000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-16] ingest | Discord link ingest X digest reviewed, no sources saved
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,720 share rows.
- Scanned 3 total #chat/#tw messages newer than `2026-07-16T14:53:01.389000000Z`; all 3 contained URLs.
- Extracted 30 URL mentions / 30 normalized unique URLs. Resolved 3 shortlinks: Kimi official page, an Asahi article, and an X/Google post.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Link-only/skipped: 30.
- Extraction blocker: `https://www.kimi.com/` matched the AI-agent/knowledge-work interest profile but Defuddle returned an empty markdown body and fallback HTML only exposed SEO metadata/title, so it stayed link-only rather than raw-ingesting a thin app shell.
- Other skips were X-only frontier-model/product/geopolitics/general-news digest links or general/local news below the strict wiki threshold.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T16:56:46.351000000Z`.
- Rubric unchanged: this repeated the existing rule that model/product chatter needs a durable, substantive source before raw/wiki ingest.
## [2026-07-16] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,728 share rows.
- Scanned 0 total #chat/#tw messages newer than `2026-07-16T16:56:46.351000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-16T16:56:46.351000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-16] ingest | Discord link ingest X digest reviewed, no sources saved
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,744 share rows.
- Scanned 4 total new #chat/#tw messages newer than `2026-07-16T16:56:46.351000000Z`; all 4 were #tw digest messages.
- Extracted 48 URL mentions / 40 normalized unique URLs. All were X/Twitter links or shortened `t.co` links embedded in the digest.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Link-only/skipped: 40. Extraction errors: 0.
- Notable skipped candidates: VS Code agent feature teaser (`x.com/code/...`) and Kimi K3 model-race discussion. Both stayed link-only because the run exposed no durable primary article/docs/repo source.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T19:55:43.405000000Z`.
- Rubric unchanged: repeated X-only digest evidence supports the existing strict rule that product/model chatter needs durable source material before raw/wiki ingest.
## [2026-07-16] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,818 share rows.
- Scanned 0 total #chat/#tw messages newer than `2026-07-16T19:55:43.405000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-16T19:55:43.405000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-16] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,846 share rows.
- Scanned 0 total #chat/#tw messages newer than `2026-07-16T19:55:43.405000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-16T19:55:43.405000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-16] ingest | Discord link ingest X digest reviewed, no sources saved
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,877 share rows.
- Scanned 3 total new #chat/#tw messages newer than `2026-07-16T19:55:43.405000000Z`; all 3 were #tw digest messages.
- Extracted 44 URL mentions / 39 normalized unique URLs. All were X/Twitter links or shortened `t.co` links embedded in the digest.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Link-only/skipped: 39. Extraction errors: 0.
- Notable skipped candidates: Kimi K3 benchmark/model-race links, Starship/SpaceX launch-documentary links, SAORI audio-generation demo, GoDaddy Developer Platform API mention, Unity/Fable workflow tweets, and geopolitics/markets/general-news items.
- No source met the strict raw-ingest threshold because the run exposed no substantive primary article/docs/repo source; shortlinks stayed link-only rather than fabricating resolved destinations.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T22:56:12.783000000Z`.
- Rubric unchanged: repeated X digest/product/model evidence supports the existing rule that chatter needs durable source material before raw/wiki ingest.
## [2026-07-17] ingest | Legal document open data from Discord links
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 178,201 share rows; archive now reports 167,271 messages.
- Scanned 4 total new #chat/#tw messages newer than `2026-07-16T22:56:12.783000000Z`; 3 messages contained URLs. One Discord CDN JPG attachment was excluded as media-only.
- Extracted 3 URL mentions / 3 normalized unique URLs.
- Raw sources saved: `raw/articles/ndl-akoma-ntoso-legal-xml-schema-2014-2026.md`, `raw/articles/e-legislation-japanese-law-akn-search-2026.md`.
- Wiki page created: `concepts/legal-document-open-data.md`.
- Wiki pages updated: `concepts/extensible-data-file-formats.md`, `index.md`.
- Link-only/skipped: 1 X-only URL plus 1 Discord CDN image attachment. Extraction blockers: `web_extract` could not fetch the NDL PDF because the configured backend is search-only; fallback direct `curl` + temporary Nix `poppler-utils`/`pdftotext` succeeded.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-16T23:43:06.946000000Z`.
- Rubric unchanged: legal/legislative open-data links fit the existing civic-tech + data-format + knowledge-infrastructure rubric.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,947 share rows; archive now reports 167,290 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-16T23:43:06.946000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-16T23:43:06.946000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Slack agentic testing E2E stack
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 173,962 share rows; archive now reports 167,297 messages.
- Scanned 1 total new #chat/#tw message newer than `2026-07-16T23:43:06.946000000Z`; 1 message contained a URL.
- Extracted 1 URL mention / 1 normalized unique URL.
- Source saved: `raw/articles/slack-agentic-testing-e2e-stack-2026.md` — score 3, Slack Engineering's 200+ run comparison of Playwright MCP, Playwright CLI, and generated Playwright tests for agentic E2E testing.
- Wiki pages updated: `concepts/e2e-coverage-metrics.md`, `concepts/agent-harness-engineering.md`.
- Wiki pages created: 0. Link-only/skipped: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T01:57:41.058000000Z`.
- Rubric unchanged: this reinforces the existing interest-profile rule for implementation-derived quality metrics and concrete agent harness reliability/cost evidence.
## [2026-07-17] ingest | Social media age assurance and child online safety
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded; `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,062 share rows; archive now reports 167,346 messages.
- Scanned 5 total new #chat/#tw messages newer than `2026-07-17T01:57:41.058000000Z`; 2 messages contained URLs.
- Extracted 3 URL mentions / 3 normalized unique URLs.
- Sources saved: `raw/articles/unicef-social-media-age-restrictions-child-rights-2026.md`, `raw/articles/yonhap-korea-social-media-age-restrictions-2026.md`.
- Wiki page created: `concepts/social-media-age-assurance.md`.
- Wiki pages updated: `concepts/data-protection-and-expression.md`, `index.md`.
- Link-only/skipped: 1 — OECD social-media age-restrictions blog was blocked by Cloudflare/403 for defuddle/direct fetch, so it was recorded as link-only rather than fabricated. Extraction errors: 1.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T03:42:01.292000000Z`.
- Rubric unchanged: the keep decision fits existing public-interest, privacy, data-protection, and inclusive-design rules; this run also reinforces that KYC/privacy concerns are a high-signal discriminator for age-assurance links.
## [2026-07-17] ingest | Discord link ingest X-only status skipped
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,087 share rows; archive now reports 167,358 messages.
- Scanned 1 total new #chat/#tw message newer than `2026-07-17T03:42:01.292000000Z`; it contained 1 URL.
- Extracted 1 URL mention / 1 normalized unique URL.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Link-only/skipped: 1 — `https://x.com/i/status/2077949194439102783`, because the Discord message contained only an X status URL and exposed no durable primary source/article/repo/docs page.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T03:53:46.300000000Z`.
- Rubric unchanged: repeated X-only evidence supports the existing strict rule that social posts need durable reusable source material before raw/wiki ingest.
## [2026-07-17] ingest | Discord link ingest Kimi K3/X digest reviewed, no sources saved
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,125 share rows; archive now reports 167,375 messages.
- Scanned 4 total new #chat/#tw messages newer than `2026-07-17T03:53:46.300000000Z`; all 4 were #tw digest fragments and all 4 contained URLs.
- Extracted 44 URL mentions / 38 normalized unique URLs.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0.
- Link-only/skipped: 38. Notable candidates were Kimi K3/Artificial Analysis/ITmedia/GIGAZINE links, a Claude Code tmux-continuation skill tweet, and GOROman’s browser event-simulation game, but the run exposed only X/Twitter URLs or unresolved `t.co` shortlinks rather than durable primary source/article/repo/docs pages.
- Extraction/search blockers: 7 `t.co` shortlinks could not be safely resolved in this cron environment: the terminal shortlink request was held for approval, `web_extract` is configured to a search-only backend, and `web_search` could not reach SearXNG.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T04:58:03.674000000Z`.
- Rubric unchanged: repeated X/unresolved-shortlink digest evidence supports the existing strict rule that product/model chatter needs durable reusable source material before raw/wiki ingest.
## [2026-07-17] ingest | Discord link ingest no new URLs
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,131 share rows; archive now reports 167,378 messages.
- Scanned 1 total new #chat/#tw message newer than `2026-07-17T04:58:03.674000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T06:47:22.829000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest X/Guardian live links skipped
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,155 share rows; archive now reports 167,389 messages.
- Scanned 3 total new #chat/#tw messages newer than `2026-07-17T06:47:22.829000000Z`; 2 messages contained URLs.
- Extracted 2 URL mentions / 2 normalized unique URLs.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0.
- Link-only/skipped: 2 — one X status only, and one Guardian live politics update. The Guardian source extracted enough text to classify, but its live-news format and shallow one-off Discord context did not meet the strict raw-ingest threshold.
- Extraction/search blockers: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T07:48:21.147000000Z`.
- Rubric unchanged: evidence matched existing strict rules for X-only and shallow/live political-news links.
## [2026-07-17] ingest | Discord link ingest X digest reviewed, no sources saved
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,207 share rows; archive now reports 167,414 messages.
- Scanned 4 total new #chat/#tw messages newer than `2026-07-17T07:48:21.147000000Z`; 3 messages contained URLs.
- Extracted 35 URL mentions / 32 normalized unique URLs.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0.
- Link-only/skipped: 32. Notable candidates included 国旗損壊罪 / expression-freedom debate, Kimi K3, ChatGPT/Codex naming confusion, NVIDIA/Japan AI support, Meta model-bias discussion, a Vim game, 分身ロボットスナック, and AI文章の「AI臭さ」; all exposed links were X status URLs, not durable primary source/article/repo/docs pages.
- Extraction/search blockers: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T07:55:43.542000000Z`.
- Rubric unchanged: repeated X-only digest evidence supports the existing strict rule that product/model/public-news chatter needs durable reusable source material before raw/wiki ingest.
## [2026-07-17] ingest | Discord-discovered Splink and record linkage sources
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,226 share rows; archive now reports 167,423 messages.
- Scanned 3 total new #chat/#tw messages newer than `2026-07-17T07:55:43.542000000Z`; all 3 contained URLs.
- Extracted 3 URL mentions / 3 normalized unique URLs.
- Raw sources saved:
- `raw/articles/splink-probabilistic-record-linkage-docs-2026.md` — score 4, Splink documentation for scalable probabilistic record linkage and public-sector use cases.
- `raw/articles/splink-github-repo-2026.md` — score 2, GitHub README / implementation reference for the same tool.
- `raw/articles/nii-record-linkage-research-trends-2004.md` — score 4, NII review PDF on record linkage, deduplication, data cleaning, blocking, and Fellegi-Sunter.
- Wiki pages created:
- `entities/splink.md`
- `concepts/record-linkage-data-cleaning.md`
- Navigation updated: `index.md` total pages 58.
- Verification: new raw/page files exist; new index entries are present; wikilinks in the two new pages resolve.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T09:50:31.677000000Z`.
- Rubric note: public-sector data integration / record linkage deserves high score when linked to durable source docs or papers, especially when quality, explainability, civic service delivery, and data-protection risks are explicit.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,246 share rows; archive now reports 167,433 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-17T09:50:31.677000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-17T09:50:31.677000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest skipped X digest batch
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,285 share rows; archive now reports 167,451 messages.
- Scanned 4 total new #chat/#tw messages newer than `2026-07-17T09:50:31.677000000Z`; all 4 contained URLs.
- Extracted 54 URL mentions / 49 normalized unique URLs.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0.
- Link-only/skipped: 49. The batch was mostly X/Twitter status links and 7 `t.co` shortlinks; unattended expansion of shortlinks was blocked by the command guard, so no linked page body was fetched or fabricated.
- Notable link-only candidates: Kimi K3 CNBC/Bloomberg coverage, Even Realities CEO/interview, Decoy Font, AWS billing bug, and AI-writing-style commentary. All were kept below raw-ingest threshold because durable primary pages were not safely extractable in this run.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T10:58:02.263000000Z`.
- Rubric unchanged: repeated X-only digest evidence supports the existing strict rule that product/model/news chatter needs durable reusable source material before raw/wiki ingest.
## [2026-07-17] ingest | Discord link ingest no discovery URLs
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 176,847 share rows; archive now reports 167,480 messages.
- Scanned 13 total new #chat/#tw messages newer than `2026-07-17T10:58:02.263000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- The new messages were Hermes cron/debug conversation, including an August report-progress automation fix and an already-handled record-linkage duplicate-ingest check; no linked source pages were present.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T12:33:12.115000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,403 share rows; archive now reports 167,495 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-17T12:33:12.115000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-17T12:33:12.115000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,479 share rows; archive now reports 167,533 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-17T12:33:12.115000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-17T12:33:12.115000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,535 share rows; archive now reports 167,561 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-17T12:33:12.115000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-17T12:33:12.115000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest X digest reviewed, no sources saved
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,590 share rows; archive now reports 167,588 messages.
- Scanned 3 total new #chat/#tw messages newer than `2026-07-17T12:33:12.115000000Z`; all 3 were #tw digest fragments and all 3 contained URLs.
- Extracted 42 URL mentions / 37 normalized unique URLs.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0.
- Link-only/skipped: 37. Notable link-only candidates: Kimi K3 / Moonshot model-race posts, Meta-Anthropic compute-leasing news, Apple/OpenAI legal-letter news, hardmaru “Diffusing Blame” paper mention, mizchi AI-writing rhythm-analysis thread, and a memory-palace interactive work. They stayed below raw-ingest threshold because the run exposed X/Twitter status URLs or unresolved `t.co` news shortlinks, not durable primary source/article/repo/docs pages.
- Extraction/search blockers: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T16:57:08.901000000Z`.
- Rubric unchanged: repeated X-only digest evidence supports the existing strict rule that model/product/research chatter needs durable reusable source material before raw/wiki ingest.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,610 share rows; archive now reports 167,598 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-17T16:57:08.901000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-17T16:57:08.901000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,614 share rows; archive now reports 167,600 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-17T16:57:08.901000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-17T16:57:08.901000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest saved WordPress security release raw
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,629 share rows; archive now reports 167,606 messages.
- Scanned 3 total new #chat/#tw messages newer than `2026-07-17T16:57:08.901000000Z`; all were #tw digest fragments with URLs.
- Extracted 41 normalized unique URL mentions. Expanded `t.co` links where possible.
- Raw articles saved: 1 — `raw/articles/wordpress-7-0-2-security-release-2026-07-17.md` from `https://wordpress.org/news/2026/07/wordpress-7-0-2-release/`.
- Wiki pages created: 0. Existing wiki pages updated: 0. The WordPress release was scored 2/raw-only: concrete security/CVE material, but not enough to update durable AI/agent/security pages.
- Link-only/skipped: 40. Notable link-only items: Palantir AIP Inspect YouTube demo, Kimi satellite tracker interactive, WSJTech SpaceX/DoD AI datacenter link blocked by 401 during redirect expansion, and X/Twitter-only model/news/geopolitics digest links.
- Extraction/search blockers: 2 — `web_extract` URL extraction backend unavailable; `https://t.co/hgyT1tHREh` returned HTTP 401. `defuddle` succeeded for the saved WordPress page.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T19:55:35.400000000Z`.
- Rubric unchanged: repeated X digest evidence supports strict link-only handling unless a durable primary source can be fetched.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,645 share rows; archive now reports 167,614 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-17T19:55:35.400000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-17T19:55:35.400000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest no new messages
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,663 share rows; archive now reports 167,623 messages.
- Scanned 0 total new #chat/#tw messages newer than `2026-07-17T19:55:35.400000000Z`; found 0 URL-containing messages.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0. Extraction errors: 0.
- Automation file updated: `.automation/discord-link-ingest/state.md`; `last_message_created_at` remains `2026-07-17T19:55:35.400000000Z`.
- Rubric unchanged: no new link evidence was available.
## [2026-07-17] ingest | Discord link ingest AI benchmarks and infrastructure ownership
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,691 share rows; archive now reports 167,635 messages.
- Scanned 5 total new #chat/#tw messages newer than `2026-07-17T19:55:35.400000000Z`; all 5 contained URLs.
- Extracted 43 URL mentions, normalized/deduped to 36 unique URLs. Expanded 5 `t.co` links with read-only Python URL resolution.
- Raw articles saved: 4 -- `raw/articles/artificial-analysis-coding-agent-benchmarks-2026.md`, `raw/articles/owners-not-renters-ai-ownership-grid-2026.md`, `raw/articles/thehackernews-wp2shell-wordpress-core-rce-2026.md`, `raw/articles/cloudflare-wordpress-waf-rules-2026.md`.
- Wiki pages created: 1 -- `concepts/ai-infrastructure-ownership.md`.
- Existing wiki pages updated: 1 -- `concepts/ai-evaluation-infrastructure.md`.
- Navigation updated: `index.md` total pages 59 and concept entry for `[[ai-infrastructure-ownership]]`.
- Link-only/skipped: 32, mostly X/Twitter digest items plus two low-context YouTube links.
- Extraction/search blockers: 1 -- `web_search`/`web_extract` backend unavailable; `defuddle` succeeded for the four saved non-X sources, with a non-fatal CSS parse warning on the Cloudflare page.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-17T23:38:31.086000000Z`.
- Rubric unchanged: evidence reinforced existing strict handling of X/YouTube links while preserving durable primary sources for model evaluation and AI infrastructure ownership.
## [2026-07-18] ingest | Discord link ingest NII journal homepage skipped
- Oriented with `SCHEMA.md`, `index.md`, recent `log.md`, and `.automation/discord-link-ingest/interest-profile.md`.
- `discrawl status --json` succeeded. `discrawl --json sql` allowed the configured git share auto-update/import path to run, importing 174,711 share rows; archive now reports 167,645 messages.
- Scanned 4 total new #chat/#tw messages newer than `2026-07-17T23:38:31.086000000Z`; 1 message contained a URL.
- Extracted 1 URL mention / 1 normalized unique URL.
- Raw articles saved: 0. Wiki pages created/updated from sources: 0.
- Link-only/skipped: 1 — `https://www.nii.ac.jp/pi/` is the old framed NII “Progress in Informatics” journal homepage; Defuddle returned an empty body and direct frame fetch showed only journal/about/current-issue navigation, so it did not meet the strict raw-ingest threshold.
- Extraction/search blockers: 1 — `web_extract` backend is search-only; fallback direct Python fetch was enough to classify the page as link-only.
- Automation file updated: `.automation/discord-link-ingest/state.md` with `last_message_created_at: 2026-07-18T00:25:36.573000000Z`.
- Rubric unchanged: this confirms the existing rule that a durable academic/institutional page still needs substantive reusable body content, not just a low-context homepage, before raw/wiki ingest.