⏪ Revert merged spike evidence

[taras]

Why

PRs #350 and #353 are evidence spikes, not production source. They were mistakenly squash-merged while collapsing their review stack. Their findings remain preserved in the PRs and issues, but their large vendored proof trees should not live on main.

What changes

Before:

main contains the complete #349 DOFS/FUSE proof and #351 Worker backend proof, including their vendored experimental dependencies.

After:

main returns exactly to the tree that existed before those two merges. PRs #350 and #353 remain the permanent reviewable evidence. PR #362 remains open and unmerged as the final stacked proof.

How it works

The branch reverts the two squash commits in reverse dependency order:

revert d96f5bf (#353)
→ revert f3f7ad5 (#350)
→ tree equals 21d26bc (pre-spike main)

Review guide

Start with: the commit-level tree comparison against 21d26bc

Then review:

1. removal of spikes/351-worker-backends/;
2. removal of spikes/349-dofs/;
3. restoration of .gitignore and deno.json; and
4. absence of every unrelated change.

Look carefully at: whether the resulting tree differs at all from pre-merge main.

What must stay true

• Spike findings remain available in PRs 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350, 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 and their linked issues.
• PR 🔬 Spike: make Worker Shell effects transactional (#357) #362 is not merged as part of this correction.
• No production or documentation work after 21d26bc is removed; no such commit landed between the two mistaken merges.

How to verify it

• git diff --check reports no whitespace errors.
• git diff --quiet 21d26bc7dd17966c4a21c151772ea5524ef2d333 -- . exits zero, proving the resulting tree exactly matches pre-spike main.
• The ordinary PR CI matrix verifies that the restored tree remains green.

Scope

Included

• Revert squash merge 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 (d96f5bf).
• Revert squash merge 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350 (f3f7ad5).

Intentionally unchanged

• PR and issue evidence for Spike: host Cloudflare DOFS directly in Deno with SQLite and FUSE #349, Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno #351 and Spike: make Worker Shell effects transactional with the workflow journal #357.
• Draft architecture PR 📝 Specify retained workflow Workspaces #358.
• Production Worker Shell issue Run shell blocks in xmd workflows #363.
• Open evidence PR 🔬 Spike: make Worker Shell effects transactional (#357) #362.

Risks and limitations

• The diff is large because it removes the vendored spike fixtures that were mistakenly added. Its resulting tree is exactly the previously green main tree.
• Recovery: the spike commits remain reachable through their merged PRs if evidence ever needs to be reconstructed locally.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and verification result.

[github-actions]

PR #364: ⏪ Revert merged spike evidence

158 files, +4 / -22864

Scope

🔴 PR has 22868 lines changed. Split into focused PRs.

🟡 22868 lines changed. PRs under 400 receive more thorough review.

🟡 158 files changed. Are all changes related?

🟡 PR mixes config and source changes.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.

[taras]

Why

PRs #350 and #353 are evidence spikes, not production source. They were mistakenly squash-merged while collapsing their review stack. Their findings remain preserved in the PRs and issues, but their large vendored proof trees should not live on main.

What changes

Before:

main contains the complete #349 DOFS/FUSE proof and #351 Worker backend proof, including their vendored experimental dependencies.

After:

main returns exactly to the tree that existed before those two merges. PRs #350 and #353 remain the permanent reviewable evidence. PR #362 remains open and unmerged as the final stacked proof.

How it works

The branch reverts the two squash commits in reverse dependency order:

revert d96f5bf (#353)
→ revert f3f7ad5 (#350)
→ tree equals 21d26bc (pre-spike main)

Review guide

Start with: the commit-level tree comparison against 21d26bc

Then review:

1. removal of spikes/351-worker-backends/;
2. removal of spikes/349-dofs/;
3. restoration of .gitignore and deno.json; and
4. absence of every unrelated change.

Look carefully at: whether the resulting tree differs at all from pre-merge main.

What must stay true

• Spike findings remain available in PRs 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350, 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 and their linked issues.
• PR 🔬 Spike: make Worker Shell effects transactional (#357) #362 is not merged as part of this correction.
• No production or documentation work after 21d26bc is removed; no such commit landed between the two mistaken merges.

How to verify it

• git diff --check reports no whitespace errors.
• git diff --quiet 21d26bc7dd17966c4a21c151772ea5524ef2d333 -- . exits zero, proving the resulting tree exactly matches pre-spike main.
• The ordinary PR CI matrix verifies that the restored tree remains green.

Scope

Included

• Revert squash merge 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 (d96f5bf).
• Revert squash merge 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350 (f3f7ad5).

Intentionally unchanged

• PR and issue evidence for Spike: host Cloudflare DOFS directly in Deno with SQLite and FUSE #349, Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno #351 and Spike: make Worker Shell effects transactional with the workflow journal #357.
• Draft architecture PR 📝 Specify retained workflow Workspaces #358.
• Production Worker Shell issue Run shell blocks in xmd workflows #363.
• Open evidence PR 🔬 Spike: make Worker Shell effects transactional (#357) #362.

Risks and limitations

• The diff is large because it removes the vendored spike fixtures that were mistakenly added. Its resulting tree is exactly the previously green main tree.
• Recovery: the spike commits remain reachable through their merged PRs if evidence ever needs to be reconstructed locally.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and verification result.

[github-actions]

PR #364: ⏪ Revert merged spike evidence

158 files, +4 / -22864

Scope

🔴 PR has 22868 lines changed. Split into focused PRs.

🟡 22868 lines changed. PRs under 400 receive more thorough review.

🟡 158 files changed. Are all changes related?

🟡 PR mixes config and source changes.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.

[taras]

Why

PRs #350 and #353 are evidence spikes, not production source. They were mistakenly squash-merged while collapsing their review stack. Their findings remain preserved in the PRs and issues, but their large vendored proof trees should not live on main.

What changes

Before:

main contains the complete #349 DOFS/FUSE proof and #351 Worker backend proof, including their vendored experimental dependencies.

After:

main returns exactly to the tree that existed before those two merges. PRs #350 and #353 remain the permanent reviewable evidence. PR #362 remains open and unmerged as the final stacked proof.

How it works

The branch reverts the two squash commits in reverse dependency order:

revert d96f5bf (#353)
→ revert f3f7ad5 (#350)
→ tree equals 21d26bc (pre-spike main)

Review guide

Start with: the commit-level tree comparison against 21d26bc

Then review:

1. removal of spikes/351-worker-backends/;
2. removal of spikes/349-dofs/;
3. restoration of .gitignore and deno.json; and
4. absence of every unrelated change.

Look carefully at: whether the resulting tree differs at all from pre-merge main.

What must stay true

• Spike findings remain available in PRs 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350, 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 and their linked issues.
• PR 🔬 Spike: make Worker Shell effects transactional (#357) #362 is not merged as part of this correction.
• No production or documentation work after 21d26bc is removed; no such commit landed between the two mistaken merges.

How to verify it

• git diff --check reports no whitespace errors.
• git diff --quiet 21d26bc7dd17966c4a21c151772ea5524ef2d333 -- . exits zero, proving the resulting tree exactly matches pre-spike main.
• The ordinary PR CI matrix verifies that the restored tree remains green.

Scope

Included

• Revert squash merge 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 (d96f5bf).
• Revert squash merge 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350 (f3f7ad5).

Intentionally unchanged

• PR and issue evidence for Spike: host Cloudflare DOFS directly in Deno with SQLite and FUSE #349, Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno #351 and Spike: make Worker Shell effects transactional with the workflow journal #357.
• Draft architecture PR 📝 Specify retained workflow Workspaces #358.
• Production Worker Shell issue Run shell blocks in xmd workflows #363.
• Open evidence PR 🔬 Spike: make Worker Shell effects transactional (#357) #362.

Risks and limitations

• The diff is large because it removes the vendored spike fixtures that were mistakenly added. Its resulting tree is exactly the previously green main tree.
• Recovery: the spike commits remain reachable through their merged PRs if evidence ever needs to be reconstructed locally.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and verification result.

[github-actions]

PR #364: ⏪ Revert merged spike evidence

158 files, +4 / -22864

Scope

🔴 PR has 22868 lines changed. Split into focused PRs.

🟡 22868 lines changed. PRs under 400 receive more thorough review.

🟡 158 files changed. Are all changes related?

🟡 PR mixes config and source changes.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.

[taras]

Why

PRs #350 and #353 are evidence spikes, not production source. They were mistakenly squash-merged while collapsing their review stack. Their findings remain preserved in the PRs and issues, but their large vendored proof trees should not live on main.

What changes

Before:

main contains the complete #349 DOFS/FUSE proof and #351 Worker backend proof, including their vendored experimental dependencies.

After:

main returns exactly to the tree that existed before those two merges. PRs #350 and #353 remain the permanent reviewable evidence. PR #362 remains open and unmerged as the final stacked proof.

How it works

The branch reverts the two squash commits in reverse dependency order:

revert d96f5bf (#353)
→ revert f3f7ad5 (#350)
→ tree equals 21d26bc (pre-spike main)

Review guide

Start with: the commit-level tree comparison against 21d26bc

Then review:

1. removal of spikes/351-worker-backends/;
2. removal of spikes/349-dofs/;
3. restoration of .gitignore and deno.json; and
4. absence of every unrelated change.

Look carefully at: whether the resulting tree differs at all from pre-merge main.

What must stay true

• Spike findings remain available in PRs 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350, 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 and their linked issues.
• PR 🔬 Spike: make Worker Shell effects transactional (#357) #362 is not merged as part of this correction.
• No production or documentation work after 21d26bc is removed; no such commit landed between the two mistaken merges.

How to verify it

• git diff --check reports no whitespace errors.
• git diff --quiet 21d26bc7dd17966c4a21c151772ea5524ef2d333 -- . exits zero, proving the resulting tree exactly matches pre-spike main.
• The ordinary PR CI matrix verifies that the restored tree remains green.

Scope

Included

• Revert squash merge 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 (d96f5bf).
• Revert squash merge 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350 (f3f7ad5).

Intentionally unchanged

• PR and issue evidence for Spike: host Cloudflare DOFS directly in Deno with SQLite and FUSE #349, Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno #351 and Spike: make Worker Shell effects transactional with the workflow journal #357.
• Draft architecture PR 📝 Specify retained workflow Workspaces #358.
• Production Worker Shell issue Run shell blocks in xmd workflows #363.
• Open evidence PR 🔬 Spike: make Worker Shell effects transactional (#357) #362.

Risks and limitations

• The diff is large because it removes the vendored spike fixtures that were mistakenly added. Its resulting tree is exactly the previously green main tree.
• Recovery: the spike commits remain reachable through their merged PRs if evidence ever needs to be reconstructed locally.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and verification result.

[github-actions]

PR #364: ⏪ Revert merged spike evidence

158 files, +4 / -22864

Scope

🔴 PR has 22868 lines changed. Split into focused PRs.

🟡 22868 lines changed. PRs under 400 receive more thorough review.

🟡 158 files changed. Are all changes related?

🟡 PR mixes config and source changes.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.

[taras]

Why

PRs #350 and #353 are evidence spikes, not production source. They were mistakenly squash-merged while collapsing their review stack. Their findings remain preserved in the PRs and issues, but their large vendored proof trees should not live on main.

What changes

Before:

main contains the complete #349 DOFS/FUSE proof and #351 Worker backend proof, including their vendored experimental dependencies.

After:

main returns exactly to the tree that existed before those two merges. PRs #350 and #353 remain the permanent reviewable evidence. PR #362 remains open and unmerged as the final stacked proof.

How it works

The branch reverts the two squash commits in reverse dependency order:

revert d96f5bf (#353)
→ revert f3f7ad5 (#350)
→ tree equals 21d26bc (pre-spike main)

Review guide

Start with: the commit-level tree comparison against 21d26bc

Then review:

1. removal of spikes/351-worker-backends/;
2. removal of spikes/349-dofs/;
3. restoration of .gitignore and deno.json; and
4. absence of every unrelated change.

Look carefully at: whether the resulting tree differs at all from pre-merge main.

What must stay true

• Spike findings remain available in PRs 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350, 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 and their linked issues.
• PR 🔬 Spike: make Worker Shell effects transactional (#357) #362 is not merged as part of this correction.
• No production or documentation work after 21d26bc is removed; no such commit landed between the two mistaken merges.

How to verify it

• git diff --check reports no whitespace errors.
• git diff --quiet 21d26bc7dd17966c4a21c151772ea5524ef2d333 -- . exits zero, proving the resulting tree exactly matches pre-spike main.
• The ordinary PR CI matrix verifies that the restored tree remains green.

Scope

Included

• Revert squash merge 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 (d96f5bf).
• Revert squash merge 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350 (f3f7ad5).

Intentionally unchanged

• PR and issue evidence for Spike: host Cloudflare DOFS directly in Deno with SQLite and FUSE #349, Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno #351 and Spike: make Worker Shell effects transactional with the workflow journal #357.
• Draft architecture PR 📝 Specify retained workflow Workspaces #358.
• Production Worker Shell issue Run shell blocks in xmd workflows #363.
• Open evidence PR 🔬 Spike: make Worker Shell effects transactional (#357) #362.

Risks and limitations

• The diff is large because it removes the vendored spike fixtures that were mistakenly added. Its resulting tree is exactly the previously green main tree.
• Recovery: the spike commits remain reachable through their merged PRs if evidence ever needs to be reconstructed locally.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and verification result.

[github-actions]

PR #364: ⏪ Revert merged spike evidence

158 files, +4 / -22864

Scope

🔴 PR has 22868 lines changed. Split into focused PRs.

🟡 22868 lines changed. PRs under 400 receive more thorough review.

🟡 158 files changed. Are all changes related?

🟡 PR mixes config and source changes.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.

[taras]

Why

PRs #350 and #353 are evidence spikes, not production source. They were mistakenly squash-merged while collapsing their review stack. Their findings remain preserved in the PRs and issues, but their large vendored proof trees should not live on main.

What changes

Before:

main contains the complete #349 DOFS/FUSE proof and #351 Worker backend proof, including their vendored experimental dependencies.

After:

main returns exactly to the tree that existed before those two merges. PRs #350 and #353 remain the permanent reviewable evidence. PR #362 remains open and unmerged as the final stacked proof.

How it works

The branch reverts the two squash commits in reverse dependency order:

revert d96f5bf (#353)
→ revert f3f7ad5 (#350)
→ tree equals 21d26bc (pre-spike main)

Review guide

Start with: the commit-level tree comparison against 21d26bc

Then review:

1. removal of spikes/351-worker-backends/;
2. removal of spikes/349-dofs/;
3. restoration of .gitignore and deno.json; and
4. absence of every unrelated change.

Look carefully at: whether the resulting tree differs at all from pre-merge main.

What must stay true

• Spike findings remain available in PRs 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350, 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 and their linked issues.
• PR 🔬 Spike: make Worker Shell effects transactional (#357) #362 is not merged as part of this correction.
• No production or documentation work after 21d26bc is removed; no such commit landed between the two mistaken merges.

How to verify it

• git diff --check reports no whitespace errors.
• git diff --quiet 21d26bc7dd17966c4a21c151772ea5524ef2d333 -- . exits zero, proving the resulting tree exactly matches pre-spike main.
• The ordinary PR CI matrix verifies that the restored tree remains green.

Scope

Included

• Revert squash merge 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 (d96f5bf).
• Revert squash merge 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350 (f3f7ad5).

Intentionally unchanged

• PR and issue evidence for Spike: host Cloudflare DOFS directly in Deno with SQLite and FUSE #349, Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno #351 and Spike: make Worker Shell effects transactional with the workflow journal #357.
• Draft architecture PR 📝 Specify retained workflow Workspaces #358.
• Production Worker Shell issue Run shell blocks in xmd workflows #363.
• Open evidence PR 🔬 Spike: make Worker Shell effects transactional (#357) #362.

Risks and limitations

• The diff is large because it removes the vendored spike fixtures that were mistakenly added. Its resulting tree is exactly the previously green main tree.
• Recovery: the spike commits remain reachable through their merged PRs if evidence ever needs to be reconstructed locally.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and verification result.

[github-actions]

PR #364: ⏪ Revert merged spike evidence

158 files, +4 / -22864

Scope

🔴 PR has 22868 lines changed. Split into focused PRs.

🟡 22868 lines changed. PRs under 400 receive more thorough review.

🟡 158 files changed. Are all changes related?

🟡 PR mixes config and source changes.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.

[taras]

Why

PRs #350 and #353 are evidence spikes, not production source. They were mistakenly squash-merged while collapsing their review stack. Their findings remain preserved in the PRs and issues, but their large vendored proof trees should not live on main.

What changes

Before:

main contains the complete #349 DOFS/FUSE proof and #351 Worker backend proof, including their vendored experimental dependencies.

After:

main returns exactly to the tree that existed before those two merges. PRs #350 and #353 remain the permanent reviewable evidence. PR #362 remains open and unmerged as the final stacked proof.

How it works

The branch reverts the two squash commits in reverse dependency order:

revert d96f5bf (#353)
→ revert f3f7ad5 (#350)
→ tree equals 21d26bc (pre-spike main)

Review guide

Start with: the commit-level tree comparison against 21d26bc

Then review:

1. removal of spikes/351-worker-backends/;
2. removal of spikes/349-dofs/;
3. restoration of .gitignore and deno.json; and
4. absence of every unrelated change.

Look carefully at: whether the resulting tree differs at all from pre-merge main.

What must stay true

• Spike findings remain available in PRs 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350, 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 and their linked issues.
• PR 🔬 Spike: make Worker Shell effects transactional (#357) #362 is not merged as part of this correction.
• No production or documentation work after 21d26bc is removed; no such commit landed between the two mistaken merges.

How to verify it

• git diff --check reports no whitespace errors.
• git diff --quiet 21d26bc7dd17966c4a21c151772ea5524ef2d333 -- . exits zero, proving the resulting tree exactly matches pre-spike main.
• The ordinary PR CI matrix verifies that the restored tree remains green.

Scope

Included

• Revert squash merge 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 (d96f5bf).
• Revert squash merge 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350 (f3f7ad5).

Intentionally unchanged

• PR and issue evidence for Spike: host Cloudflare DOFS directly in Deno with SQLite and FUSE #349, Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno #351 and Spike: make Worker Shell effects transactional with the workflow journal #357.
• Draft architecture PR 📝 Specify retained workflow Workspaces #358.
• Production Worker Shell issue Run shell blocks in xmd workflows #363.
• Open evidence PR 🔬 Spike: make Worker Shell effects transactional (#357) #362.

Risks and limitations

• The diff is large because it removes the vendored spike fixtures that were mistakenly added. Its resulting tree is exactly the previously green main tree.
• Recovery: the spike commits remain reachable through their merged PRs if evidence ever needs to be reconstructed locally.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and verification result.

[github-actions]

PR #364: ⏪ Revert merged spike evidence

158 files, +4 / -22864

Scope

🔴 PR has 22868 lines changed. Split into focused PRs.

🟡 22868 lines changed. PRs under 400 receive more thorough review.

🟡 158 files changed. Are all changes related?

🟡 PR mixes config and source changes.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.

[taras]

Why

PRs #350 and #353 are evidence spikes, not production source. They were mistakenly squash-merged while collapsing their review stack. Their findings remain preserved in the PRs and issues, but their large vendored proof trees should not live on main.

What changes

Before:

main contains the complete #349 DOFS/FUSE proof and #351 Worker backend proof, including their vendored experimental dependencies.

After:

main returns exactly to the tree that existed before those two merges. PRs #350 and #353 remain the permanent reviewable evidence. PR #362 remains open and unmerged as the final stacked proof.

How it works

The branch reverts the two squash commits in reverse dependency order:

revert d96f5bf (#353)
→ revert f3f7ad5 (#350)
→ tree equals 21d26bc (pre-spike main)

Review guide

Start with: the commit-level tree comparison against 21d26bc

Then review:

1. removal of spikes/351-worker-backends/;
2. removal of spikes/349-dofs/;
3. restoration of .gitignore and deno.json; and
4. absence of every unrelated change.

Look carefully at: whether the resulting tree differs at all from pre-merge main.

What must stay true

• Spike findings remain available in PRs 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350, 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 and their linked issues.
• PR 🔬 Spike: make Worker Shell effects transactional (#357) #362 is not merged as part of this correction.
• No production or documentation work after 21d26bc is removed; no such commit landed between the two mistaken merges.

How to verify it

• git diff --check reports no whitespace errors.
• git diff --quiet 21d26bc7dd17966c4a21c151772ea5524ef2d333 -- . exits zero, proving the resulting tree exactly matches pre-spike main.
• The ordinary PR CI matrix verifies that the restored tree remains green.

Scope

Included

• Revert squash merge 🔬 Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno (#351) #353 (d96f5bf).
• Revert squash merge 🔬 Spike: host Cloudflare DOFS directly in Deno with SQLite (#349) #350 (f3f7ad5).

Intentionally unchanged

• PR and issue evidence for Spike: host Cloudflare DOFS directly in Deno with SQLite and FUSE #349, Spike: run Cloudflare Worker Shell and Worker JavaScript natively in Deno #351 and Spike: make Worker Shell effects transactional with the workflow journal #357.
• Draft architecture PR 📝 Specify retained workflow Workspaces #358.
• Production Worker Shell issue Run shell blocks in xmd workflows #363.
• Open evidence PR 🔬 Spike: make Worker Shell effects transactional (#357) #362.

Risks and limitations

• The diff is large because it removes the vendored spike fixtures that were mistakenly added. Its resulting tree is exactly the previously green main tree.
• Recovery: the spike commits remain reachable through their merged PRs if evidence ever needs to be reconstructed locally.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and verification result.

[github-actions]

PR #364: ⏪ Revert merged spike evidence

158 files, +4 / -22864

Scope

🔴 PR has 22868 lines changed. Split into focused PRs.

🟡 22868 lines changed. PRs under 400 receive more thorough review.

🟡 158 files changed. Are all changes related?

🟡 PR mixes config and source changes.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.