Skip to content

fix(queen): index PASSPORT inside PROJECT and read L0 from the constitution - #1162

Merged
dmitrii-f-t27 merged 1 commit into
gHashTag:mainfrom
dmitrii-f-t27:fix/queen-index-and-l0
Sep 24, 2026
Merged

dmitrii-f-t27 merged 1 commit into
gHashTag:mainfrom
dmitrii-f-t27:fix/queen-index-and-l0

Conversation

@dmitrii-f-t27

Copy link
Copy Markdown
Collaborator

The Queen map on app.t27.ai (and the #/specs catalog) still draws the atlas built on 2026-09-21 12:24 UTC: dmitrii-f-t27/trinity-memory shows 0 issues there and has 17 open today. The nightly t27-world-scan has failed every night since 2026-09-17; #1154 fixed the zoom ceiling and the push lease, and the manual scan run 36025169229 on main then stopped at check:agents.

What was left is drift in the site, not in the data:

  • check:agents — PASSPORT moved inside PROJECT on 2026-09-21 (PROJECT_VIEWS), but the homepage index (ModulesBlock) and its contract still knew three groups, so PASSPORT had no card anywhere and the view partition no longer covered every view. The index now draws a fourth group, the record inside PROJECT, built like the board group; the contract checks four groups, the project family, and the rail length with it.
  • check:docs — the constitution adopted L0 PURPOSE on 2026-09-23 and states it sits outside the L1 > … > L7 ordering. The contract accepts L0 ahead of L1–L7 (which must still follow in order), and the law figure draws L0 as a dashed goal band above the pyramid instead of a rung, so it never joins the > chain.

Verified: a local run of the scan's whole pipeline on this branch, followed by all twelve catalog gates, passes on fresh data — 12 worlds, 1577 specs, 1237 issues, trinity-memory 17 issues / 93 specs. After merge, the next scan (manual dispatch or the 03:17 UTC schedule) opens its data PR on bot/t27-world-scan; Railway rebuilds app.t27.ai from main once that lands.

{
  "version": 1,
  "head_sha": "c61f2c2c2ec8fea024a622411eafc17a119c5210",
  "summary": "The Queen map on app.t27.ai still draws the 2026-09-21 atlas because the nightly t27-world-scan fails its catalog gates. After #1154 the two remaining failures are drift in the site, not in the data: PASSPORT moved inside PROJECT without the homepage index learning a fourth group, and the constitution adopted L0 PURPOSE while the docs contract still expects exactly L1-L7. This fixes both so the scan can publish a fresh atlas.",
  "changes": [
    "src/components/ModulesBlock.tsx: a fourth index group, the record inside PROJECT, built like the board group (isProjectRecord over PROJECT_VIEWS); lede and headings in en and ru count it",
    "qa/agents-spec-contract.mjs: four index groups, the project family in the view partition, only PROJECT on both rail and family, rail length counts the record folded into PROJECT",
    "qa/docs-spec-contract.mjs: L0 PURPOSE may open the law table; L1-L7 must still follow in order",
    "src/components/SystemDocsFigures.tsx: the law figure draws L0 as a dashed goal band above the pyramid, outside the '>' priority chain, as the constitution states"
  ],
  "tests": [
    {
      "command": "local run of the t27-world-scan pipeline (sync, discover scan/vendor, core index, skills, agents, docs, atlas scan/build)",
      "result": "Completed on fresh data: 12 worlds, 1577 specs, 1237 issues; dmitrii-f-t27/trinity-memory backlog 17 issues, 93 specs",
      "status": "passed",
      "evidence": "Local run on this branch, 2026-09-24 13:16-13:57 -03, GH token read-only use"
    },
    {
      "command": "npm run check:discovery, check:shared-core, check:spec-catalog, check:queen-atlas, check:queen-catalog, check:queen-continuous, check:queen-spec-mirrors, check:skills-catalog, check:crons-catalog, check:agents, check:tools, check:docs",
      "result": "All twelve pass on the freshly scanned tree",
      "status": "passed",
      "evidence": "Same gates as the scan's Catalog gates step; on main the manual scan run 36025169229 failed at check:agents on passport"
    },
    {
      "command": "npx eslint src/components/ModulesBlock.tsx src/components/SystemDocsFigures.tsx",
      "result": "No lint findings in either file",
      "status": "passed",
      "evidence": "Local run on this branch, eslint from apps/website"
    },
    {
      "command": "npx tsc --noEmit -p tsconfig.app.json",
      "result": "179 errors both with and without this change, none in the two changed components",
      "status": "passed",
      "evidence": "Compared against a stash of the change; the errors are in unrelated files such as src/pages/TrinityCanvas.tsx"
    },
    {
      "command": "python3 scripts/pr_blog_report.py validate --event event.json --output out",
      "result": "This PR body validates against the work report contract",
      "status": "passed",
      "evidence": "Local run of the script from main with a synthetic event at this head"
    }
  ],
  "limitations": [
    "Only the site code changes here; the refreshed catalog data arrives through the scan's own pull request (bot/t27-world-scan) after this merges",
    "deploy-site.yml (t27.ai on GitHub Pages) fails on a missing token secret; app.t27.ai is built by Railway from main and does not depend on it"
  ],
  "tags": [
    "queen",
    "catalog",
    "ci",
    "docs"
  ],
  "blog": {
    "title": "Why the Queen map stopped at September 21, and the two drifts that held it there",
    "summary": "The public Queen map is a static atlas rebuilt by a nightly scan. The scan had failed every night since September 17, first on a push lease and a zoom ceiling, then on two contracts that no longer matched the site: a module folded into PROJECT and a new purpose law in the constitution.",
    "outline": [
      "The Queen map and spec catalog render a static atlas that a nightly job rebuilds from every repository that writes .t27 specs.",
      "The job had failed since September 17, so newer issues and specs, such as the seventeen open issues in trinity-memory, never reached the map.",
      "After the zoom and push fix in #1154, the scan stopped on PASSPORT, which had moved inside PROJECT without the homepage index learning a fourth group.",
      "The docs contract still expected exactly seven laws after the constitution adopted L0 PURPOSE, which it places outside the priority ordering.",
      "With both fixed, the full pipeline and all twelve catalog gates pass on fresh data, and the next scan can publish the refreshed atlas."
    ]
  }
}

🤖 Generated with Claude Code

…tution

The nightly t27-world-scan has not refreshed the catalog since 2026-09-16,
so app.t27.ai still draws the 2026-09-21 atlas (trinity-memory: 0 issues,
now 17). After gHashTag#1154 the next two catalog gates failed on drift, not on data:

- check:agents: PASSPORT moved inside PROJECT on 2026-09-21 (PROJECT_VIEWS),
  but the homepage index and the contract still knew three groups, so
  PASSPORT had no card anywhere and the view partition missed it. The index
  now draws a fourth group, the record inside PROJECT; the contract checks
  four groups, the project family and the rail length with it.
- check:docs: the constitution adopted L0 PURPOSE on 2026-09-23 and says it
  sits outside the L1 > ... > L7 ordering. The contract accepts L0 before
  L1-L7; the law figure draws L0 as a dashed goal band above the pyramid
  instead of a rung, so it never joins the '>' chain.

With these, a local run of the scan's pipeline and all twelve catalog gates
pass on fresh data (12 worlds, 1577 specs, 1237 issues; trinity-memory 17).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@dmitrii-f-t27
dmitrii-f-t27 merged commit ed76c99 into gHashTag:main Sep 24, 2026
21 of 27 checks passed
github-actions Bot added a commit that referenced this pull request Sep 24, 2026
fix(queen): index PASSPORT inside PROJECT and read L0 from the constitution (#1162)

The last two catalog gates holding the nightly t27-world-scan: the homepage index learns the PROJECT family PASSPORT moved into on 2026-09-21, and the docs contract accepts L0 PURPOSE (adopted 2026-09-23, outside the L1-L7 ordering), which the law figure draws as a goal band. A local run of the whole scan pipeline and all twelve catalog gates passes on fresh data.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant