feat(mcp): validate .loopover.yml offline by extracting buildFocusManifestValidation into @loopover/engine

[davion-knight]

What & why

loopover_validate_config's result builder, buildFocusManifestValidation, lived app-only in src/services/focus-manifest-validation.ts, so the local stdio server had to apiPost to /v1/validate/focus-manifest to validate a manifest — unusable offline. Its parsing core (parseFocusManifestContent) was already engine-side, but the result builder and its unknownTopLevelWarnings dependency (src/selfhost/config-lint.ts) were not.

This extracts both into @loopover/engine, following the established extraction pattern (a re-export shim left at each original src/ location), and rewires the local handler to compute in-process.

Changes

• New packages/loopover-engine/src/focus-manifest-validation.ts and packages/loopover-engine/src/config-lint.ts — the extracted builders, importing the manifest core engine-relative.
• Shims at src/services/focus-manifest-validation.ts and src/selfhost/config-lint.ts re-export the engine surface, so the remote server (src/mcp/server.ts) and API (src/api/routes.ts) — and every existing config-lint caller — keep working unchanged (no behavior change).
• Exports added to packages/loopover-engine/package.json (./config-lint, ./focus-manifest-validation) and the barrel.
• Local handler loopover_validate_config now calls buildFocusManifestValidation(input) directly instead of apiPost — fully offline.

The extraction necessarily brings unknownTopLevelWarnings (and its config-lint module) engine-side too, since the engine cannot import from src/; config-lint is self-contained (only yaml + engine symbols), so it moves cleanly with a shim for its other callers.

Tests

• All existing focus-manifest-validation (13), selfhost-config-lint (22), loopover-config-lint-script (12), and config-templates tests pass against the new location.
• Adds test/unit/mcp-cli-validate-config-offline.test.ts — drives the real stdio loopover_validate_config with a black-holed API URL (http://127.0.0.1:1) to prove it validates in-process (well-formed → ok, unknown field → warn, unparseable → error); a residual round-trip would fail these.
• Both new engine files are at 100% line + branch coverage via the migrated tests.

Closes #6269

[superagent-security]

Superagent didn't find any vulnerabilities or security issues in this PR.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 95.61%. Comparing base (a2d8306) to head (787d1cd).
⚠️ Report is 9 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #6425   +/-   ##
=======================================
  Coverage   95.61%   95.61%           
=======================================
  Files         600      600           
  Lines       47346    47354    +8     
  Branches    15067    15068    +1     
=======================================
+ Hits        45268    45276    +8     
  Misses       1291     1291           
  Partials      787      787           

Flag	Coverage Δ
shard-1	44.13% <5.55%> (+0.10%)	⬆️
shard-2	36.65% <5.55%> (-0.11%)	⬇️
shard-3	32.26% <83.33%> (-0.09%)	⬇️
shard-4	34.60% <5.55%> (+0.07%)	⬆️
shard-5	30.80% <82.22%> (-0.76%)	⬇️
shard-6	45.30% <38.88%> (+0.31%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
packages/loopover-engine/src/config-lint.ts	100.00% <100.00%> (ø)
...s/loopover-engine/src/focus-manifest-validation.ts	100.00% <100.00%> (ø)

... and 1 file with indirect coverage changes

[loopover-orb]

Tip

✅ LoopOver review result - approve/merge recommended

_{Review updated: 2026-07-16 07:39:40 UTC}

8 files · 1 AI reviewer · no blockers · readiness 88/100 · CI green · clean

✅ Suggested Action - Approve/Merge

• safe to merge

Review summary
This is a clean, mechanical extraction: the two builders (`buildFocusManifestValidation`, `lintManifestText`/`unknownTopLevelWarnings`) move verbatim from `src/` into `@​loopover/engine`, with shims left behind for existing callers, matching the repo's established extraction pattern (referenced as #6269 in comments). The MCP CLI handler is rewired from an `apiPost` round-trip to a direct in-process call, and the new offline test deliberately points at an unreachable API URL to prove no network round-trip occurs. The diffs are byte-for-byte moves (confirmed against full file content) with no logic changes, so behavior parity is very likely intact.

Nits — 5 non-blocking

• packages/loopover-engine/src/config-lint.ts:103 and :120 both return `null` on parse failure rather than distinguishing a genuine parse error from 'not an object', which is pre-existing behavior carried over from the move, not introduced here.
• The PR description's issue reference is `feat(mcp): extract buildFocusManifestValidation into @loopover/engine so loopover_validate_config can work offline #6269`, but the task metadata for this review references `feat(mcp): validate .loopover.yml offline by extracting buildFocusManifestValidation into @loopover/engine #6425`/`fix(config): focus-manifest validator never warns on unknown top-level fields, unlike selfhost config-lint #5929` inconsistently in comments — worth double-checking the linked issue number is accurate before merge.
• test/unit/mcp-cli-validate-config-offline.test.ts relies on a live stdio subprocess spin-up per test (beforeEach/afterEach), which is somewhat heavy for 3 assertions — consider consolidating into one test with multiple tool calls if CI time matters.
• The retained comment in config-lint.ts about `hasOwnProperty.call` vs `key in` is verbose for a moved file; fine to keep since it's unchanged content, just noting it wasn't trimmed during extraction.
• Confirm CI failures on `validate` and `validate-tests (1)` aren't related to the new package export wiring (e.g., missing `dist/config-lint.d.ts` build step) before merging, since exports were added to `package.json` but the tsconfig/build glob wasn't shown in the diff.

Decision drivers

• ✅ Code review — No blockers (1 reviewer)
• ✅ Gate result — Passing (No configured blocker found.)

Context & advisory signals — never blocks the verdict

Signal	Result	Evidence
Linked issue	✅ Linked	#6269
Related work	✅ No active overlap found	No same-issue or scoped active PR overlap found.
Change scope	❌ 8/20	High review scope from cached public metadata (1 linked issue).
Validation posture	✅ 25/25	PR body includes validation/test evidence.
Contributor workload	✅ 10/10	Author activity: 157 registered-repo PR(s), 103 merged, 0 issue(s).
Contributor context	✅ Confirmed Gittensor contributor	davion-knight; Gittensor profile; 157 PR(s), 0 issue(s).
Improvement	✅ Minor	risk: clean · value: minor · LLM: moderate

Review context

• Author: davion-knight
• Role context: outside_contributor
• Public audience mode: oss maintainer
• Lane context: Repository is configured for direct PR review.
• Public profile languages: not available
• Official Gittensor activity: 157 PR(s), 0 issue(s).
• PR-specific overlap: none found.

Contributor next steps

• Start here: Add a concise scope and risk note.
• Then work through the remaining 1 step in the Signals table above.

Signal definitions

• Related work = same linked issue, overlapping active PRs, or title/path similarity.
• Change scope = cached public metadata such as size labels, draft state, and review-burden hints.
• Validation posture = whether the PR provides enough public validation/test evidence for maintainer review.
• Contributor workload = public contributor activity and cleanup pressure, not a repo-wide quality failure.
• Contributor context = public GitHub/Gittensor identity context; non-Gittensor status is not a blocker.

🧪 Chat with LoopOver

Ask LoopOver a question about this PR directly in a comment — grounded only in the same cached, public-safe facts shown above, never a new claim.

• @loopover ask <question> answers contribution-quality Q&A with source citations and freshness.
• @loopover chat <question> answers in natural prose from cached decision-pack facts via local inference (maintainer/collaborator; read-only).
• A plain-language @loopover mention with a real question is routed to the closest matching read-only command automatically — no exact syntax required.

Full command reference: https://loopover.ai/docs/loopover-commands

🧪 Experimental — new and may change.

_{🟩 Safe / merged · 🟦 Advisory · 🟨 Held for review · 🟥 Blocked / closed}

---

💰 Earn for open-source contributions like this. Gittensor lets GitHub contributors earn for the work they already do — register to start earning →.

Checked by LoopOver, a quiet PR intelligence layer for OSS maintainers.

• Re-run LoopOver review

[loopover-orb]

LoopOver approves — the gate is satisfied and CI is green.