Skip to content

Align ecosystem tutorial links and generated notebook guidance #127

Closed

Description

@Jammy2211

Overview

Align current ecosystem tutorial links, topic descriptions and generated notebook guidance with the updated HowToLens and HowToGalaxy courses. The human approved this plan and autonomous implementation, validation and PR creation.

Plan

Fix canonical sources and routing; compare .py sources with committed notebooks; regenerate all affected consumers; validate link targets, promised topics, routing and generated parity across all52 audited checkouts. Stage claimed owners separately without overriding their claims.

Branch survey

Five available repositories have clean worktrees on feature/tutorial-link-sweep. PyAutoLens/PyAutoGalaxy remain claimed by magnification-maps-wiring; autolens_assistant by results-library-p5a-assistant (PR177). Prepare proposed patches and queue those owners.

Approved detailed plan and original request

Align ecosystem documentation with the current HowToLens and HowToGalaxy courses

Type: docs
Difficulty: large
Autonomy: supervised
Consequence: judge
Target: @PyAutoLens @PyAutoGalaxy @autolens_assistant @autogalaxy_assistant @HowToLens @howtogalaxy @autolens_workspace @autogalaxy_workspace

Original request

can we now do a sweep of all docs and make sure the way they link to HowToLens and HowToGalaxy reflects these updates, really check this for the whole ecosystem.

Audit and plan awaiting approval

Audited current default-branch tracked references across all 51 registered repositories plus the organization profile (52 checkouts). Refreshed refs, using HTTPS without changing origin configuration for two SSH-auth failures. Full inventory and raw results: /mnt/c/Users/Jammy/Workspace/Science/scratch/tutorial_link_sweep_2026-11-04/.

Initial URL validation finds 348 distinct file/URL pairs targeting missing main-branch tutorial paths across six repositories. Most are assistant-generated notebook guidance: fix the source skills and routing first, then regenerate consumers. This count excludes version-pinned pages: library docs still describe the previous 12-lesson Lens chapter and 6-lesson Galaxy chapter. Check prose-only chapter/tutorial references and pinned destinations as well as URLs.

  1. Finish a contextual mapping of old topics to current lessons across all chapters, using current HowToLens and HowToGalaxy scripts/contents as authority. Distinguish renamed, merged and removed topics; link to production guides when no current lesson teaches the promised content.
  2. Update the two libraries' docs/howto* chapter indexes and relevant overview descriptions. Make current-course links follow the current course; retain intentionally historical/version-specific records. Review release URL-bumping policy before changing pins.
  3. Repair the two assistant skill sets' tutorial links, labels and explanations, plus colab/contexts.yaml number-sensitive routing. Update relevant chapter summaries/routing in the courses. No unrelated science/API rewrite.
  4. Regenerate affected notebooks and navigation in both courses and both workspaces using the repaired assistant sources; never hand-edit generated notebook instruction cells.
  5. Repeat the whole-ecosystem audit (52 coverage receipts), validate local/remote tutorial targets and promised topics, verify routing and generated-source parity, build affected docs, run appropriate repository checks and independent review. Produce a before/after report and open linked PRs for affected repositories.

Tier: judge — merge mode: human /prm. No source edits or issues yet; this draft and review artifacts are the planning deliverables. Existing historical records remain historical. Branch/claim conflicts must be checked before worktrees are created.

Branch/claim survey

All eight affected canonical checkouts are on main and clean at audit time. Existing active claims cover PyAutoLens/PyAutoGalaxy (magnification-maps-wiring) and autolens_assistant (results-library-p5a-assistant, open PR177). Do not overlap or remove those claims. Phase implementation around available owners; queue the claimed-repository phases through start_dev unless their owners release them. No API dependency is introduced by these documentation changes, despite the generic FeatureDecision's library/API warnings.

Suggested phases: (1) canonical teaching links, topic descriptions and Colab routing in available assistant/docs owners, with claimed owners queued; (2) regenerate the four consuming notebook repositories after the relevant canonical source changes are ready; (3) repeat ecosystem-wide verification and ship documentation-only PRs with coverage/limitations recorded. One approval covers these phases; each merge remains human /prm.

Latest steering

Yes, approve the plan. Go ahead autonomously, fix the sources, regenerate the notebooks and run the checks across the ecosystem. Coordinate around the existing claims and get the PRs up. Note that notebooks are often regenerated compare to their .py files in workspaces and others, so make sure its not just that the python scripts are correct but need a regen

Activity

  1. Jammy2211 commented on Nov 4, 2026

    @Jammy2211
    CollaboratorAuthor

    Implementation/validation handoff (umbrella remains open):

    All 570 notebooks regenerated and checked against Python narrative/code, inventories, schema and intended assistant context. Regeneration restored 13 Lens notebooks whose executable content lagged their existing Python sources. Lens curated smoke: 42 scripts + 2 notebooks passed. Galaxy assistant: 84 tests passed; 221 API symbols checked with none missing. Independent review CLEAN. Two library documentation proposals pass isolated Sphinx/MyST -W builds.

    Final audit covers 52 repositories and 1,192 distinct course URLs, with zero missing active destinations in the combined five implemented worktrees plus three proposed overlays. This does not mean the queued patches are applied or published. Existing claims remain intact: PyAutoLens/PyAutoGalaxy patches queued on PyAutoLens#818; autolens_assistant patch queued on PR177. Lens consumers stay draft until that canonical dependency is applied. The Lens assistant proposal includes its COSMOS notebook metadata regeneration and passes focused tests; its full suite encounters an unchanged fixture requiring an older pinned wheel (107 passed, 1 skipped before error; separate compatible subset 294 passed).

    Review bundle: /mnt/c/Users/Jammy/Workspace/Science/scratch/tutorial_link_sweep_2026-11-04/index.html. Patches/evidence also retained in /home/jammy/Code/PyAutoLabs/.worktrees/tutorial-link-sweep/review/. Fresh Heart reason set remains the acknowledged unrelated AutoFit timeout plus stale release rehearsal. No merges or scheduled polling armed. Keep this issue and task worktree until the three claimed-owner phases are resolved; partial Galaxy merges do not complete the umbrella.

  2. Jammy2211 commented on Nov 4, 2026

    @Jammy2211
    CollaboratorAuthor

    The explicitly approved documentation-only claim exception is complete. All eight repositories now have actual task worktrees and open PRs; no patches remain queued solely because of the earlier repository claims.

    New PRs:

    Full local suites pass: PyAutoGalaxy 1,465 passed/1 skipped; PyAutoLens 1,021 passed/1 expected failure; Lens assistant 305 passed/3 skipped/1 expected failure. The assistant used its required isolated 2026.9.27.2 wheels, resolving the earlier fixture blocker without test or version-guard changes. Focused context tests, idempotent metadata generation, both documentation builds, whole-ecosystem link checks and independent review pass. The 52-repo audit covers eight actual worktrees and 1,192 course URLs with zero unresolved active destinations.

    Merge order: Galaxy assistant43 before HowToGalaxy103/autogalaxy_workspace271; Lens assistant178 before HowToLens128/autolens_workspace617. Both Lens consumers are ready for review with their canonical dependency recorded. Library docs694/819 have no API dependency on the consumers. Keep this umbrella open until all eight PRs are handled.

    Original magnification-maps-wiring and results-library-p5a-assistant claims, branches and worktrees are preserved. Committed branch changes have no path overlap with these docs changes; other worktree uncommitted changes were not accessed or modified. Current Heart is STALE only for absent release rehearsal. Seven PRs are green at the final snapshot; the new PyAutoLens documentation PR has CI running. No merges or scheduled polling armed.

    Updated review bundle: /mnt/c/Users/Jammy/Workspace/Science/scratch/tutorial_link_sweep_2026-11-04/index.html. Detailed validation is under review/docs-ship-validation in the task worktree bundle.

  3. Jammy2211 commented on Nov 4, 2026

    @Jammy2211
    CollaboratorAuthor

    Shipped — 2026-11-04

    All eight PRs are confirmed merged, with every feature branch proved contained in origin/main. All 22 CI runs and 46 jobs passed before merge; library freeze was clear.

    The ecosystem sweep is complete: 52 repositories audited, 570 notebooks regenerated/validated, and all 1,192 active course destinations resolve. Full library/assistant suites, Lens smoke, docs builds and independent review passed. Other tasks' claims and worktrees remain intact. The separate alw-optdeps-escapes task is unblocked, not completed by this work.

    Closing this umbrella and recording completion in PyAutoMind. User approved removal of the merged task worktree and generated test artifacts; the Science/scratch review bundle remains. Library pending-release labels remain; no release was performed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions