Skip to content

feat(engine): extract content-lane's pure leaf modules to loopover-engine #5775

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
loopover-orb merged 1 commit into from
Jul 14, 2026
Merged

feat(engine): extract content-lane's pure leaf modules to loopover-engine #5775

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
129 changes: 129 additions & 0 deletions packages/loopover-engine/src/review/content-lane/content-repo-spec.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,129 @@
// Modular content-repository configuration for the curated-list content lane (the awesome-claude lane and any
// self-hosted curated list). The curated-list analogue of RegistryLaneSpec (the metagraphed registry lane): a
// maintainer whose list uses different categories or a different entry-file layout parameterizes the lane via
// config instead of a gittensory code change. Defaults preserve the awesome-claude behaviour byte-for-byte.
//
// This is a LEAF module (no content-lane imports) so every consumer — scope, duplicates, source-evidence — can
// import the spec without an import cycle. Fields are added here as each consumer is migrated.
export interface ContentRepoSpec {
/** The content categories the list accepts (the first path segment under the entry root). */
categories: ReadonlySet<string>;
/** Matches one content entry file, capturing [category, slug] — e.g. /^content\/([^/]+)\/([^/]+)\.mdx$/i. */
entryPathPattern: RegExp;
/** Head-branch prefixes used by bulk maintenance automation (link-health, etc.); these legitimately edit many
* entries in one PR and are ignored, never closed. */
maintenanceBranchPrefixes: readonly string[];
/** Frontmatter fields whose edit on a MODIFIED entry is a protected close — identity / provenance /
* verification / structural / monetization + supply-chain links. */
protectedFrontmatterFields: ReadonlySet<string>;
/** URL-bearing frontmatter keys (camelCase + snake_case) normalized + compared for duplicate detection. */
urlFields: ReadonlySet<string>;
/** Generic ecosystem hosts that never make a strict/aggressive domain-only match (a shared one is at most "related"). */
domainOnlyExclusions: ReadonlySet<string>;
/** Catalog roots that legitimately back MANY entries; a shared root alone is never strict (only a shared subpath). */
multiEntryCatalogUrls: ReadonlySet<string>;
/** Scalar source-URL frontmatter fields, in extraction ORDER (the source-evidence gate reads them in sequence). */
sourceUrlFields: readonly string[];
/** Array-valued source-URL frontmatter fields (e.g. retrievalSources/sourceUrls) read as real source evidence. */
sourceUrlListFields: ReadonlySet<string>;
/** Source fields treated as distribution (download/package) rather than canonical provenance. */
distributionSourceFields: ReadonlySet<string>;
/** Hosts that classify any source URL as distribution regardless of field (package registries / artifact hosts). */
distributionSourceHosts: ReadonlySet<string>;
/** Canonical fields that anchor the close decision + block inconclusive-downgrade (the primary provenance links). */
primaryCanonicalSourceFields: ReadonlySet<string>;
}

/** The default curated-list spec — awesome-claude's categories, entry layout, and maintenance branches. */
export const AWESOME_CLAUDE_CONTENT_SPEC: ContentRepoSpec = {
categories: new Set(["agents", "collections", "commands", "guides", "hooks", "mcp", "rules", "skills", "statuslines", "tools"]),
entryPathPattern: /^content\/([^/]+)\/([^/]+)\.mdx$/i,
maintenanceBranchPrefixes: ["links/"],
// PROTECTED = identity / provenance / verification / structural / monetization + supply-chain links. The entry's
// own REFERENCE/DOCS URLs are deliberately NOT protected (those links rot + legitimately need fixing); download/
// package/affiliate URLs stay protected (supply-chain / monetization risk).
protectedFrontmatterFields: new Set([
"affiliateUrl",
"author",
"authorProfileUrl",
"category",
"claimStatus",
"claimUrl",
"dateAdded",
"disclosure",
"downloadUrl",
"importPrNumber",
"importPrUrl",
"packageUrl",
"packageVerified",
"pricingModel",
"reviewedAt",
"reviewedBy",
"reviewedPrNumber",
"slug",
"submittedAt",
"submittedBy",
"submittedByUrl",
"sourceSubmissionNumber",
"sourceSubmissionUrl",
]),
urlFields: new Set([
"documentationUrl",
"docsUrl",
"downloadUrl",
"githubUrl",
"packageUrl",
"repoUrl",
"repositoryUrl",
"sourceUrl",
"websiteUrl",
"docs_url",
"download_url",
"github_url",
"package_url",
"repo_url",
"repository_url",
"source_url",
"website_url",
]),
domainOnlyExclusions: new Set(["github.com", "npmjs.com", "pypi.org", "raw.githubusercontent.com", "registry.npmjs.org"]),
multiEntryCatalogUrls: new Set([
"https://code.claude.com/docs/en/hooks",
"https://code.claude.com/docs/en/statusline",
"https://github.com/awslabs/mcp",
"https://github.com/microsoft/mcp",
"https://github.com/modelcontextprotocol/servers",
"https://github.com/snowflake-labs/mcp",
"https://github.com/twilio-labs/mcp",
]),
sourceUrlFields: [
"documentationUrl",
"docsUrl",
"downloadUrl",
"githubUrl",
"packageUrl",
"repoUrl",
"repositoryUrl",
"sourceUrl",
"websiteUrl",
],
sourceUrlListFields: new Set(["sourceUrls", "retrievalSources"]),
distributionSourceFields: new Set(["downloadUrl", "packageUrl"]),
distributionSourceHosts: new Set([
"crates.io",
"files.pythonhosted.org",
"hub.docker.com",
"marketplace.visualstudio.com",
"mvnrepository.com",
"npmjs.com",
"packagist.org",
"pkg.go.dev",
"plugins.gradle.org",
"pypi.org",
"registry.npmjs.org",
"repo1.maven.org",
"rubygems.org",
"www.npmjs.com",
]),
primaryCanonicalSourceFields: new Set(["githubUrl", "repoUrl", "repositoryUrl", "sourceUrl"]),
};
24 changes: 24 additions & 0 deletions packages/loopover-engine/src/review/content-lane/flag.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,24 @@
// Content-lane feature flag (convergence — reviewbot→gittensory content-review port).
//
// gittensory's native code-gate reviews CODE repos. The content lane reviews CONTENT repos — a
// curated list (awesome-claude) and a registry (metagraphed) — a different domain with its own
// deterministic primitives (duplicate detection, source-evidence reachability, security scanning,
// scope classification, and metagraphed's netuid grounding). The lane is ported as native,
// self-contained gittensory modules under this directory.
//
// FLAG-GATED + DEFAULT-OFF: the lane only runs when LOOPOVER_REVIEW_CONTENT_LANE is truthy in the Env.
// Flag-off, the host never reaches these modules, so the live behavior is byte-identical. At
// cutover the host flips the flag and routes awesome-claude + metagraphed PRs through the lane.

/** Env subset the content lane reads. The full Env adds it via env.d.ts; this keeps the lane
* testable without the whole binding (pass a plain object). */
export interface ContentLaneEnv {
/** When truthy ("1"/"true"/"on"/"yes"), the content lane is enabled. Default OFF. */
LOOPOVER_REVIEW_CONTENT_LANE?: string;
}

/** Is the content lane enabled? Default OFF — only a recognized truthy flag turns it on. */
export function isContentLaneEnabled(env: ContentLaneEnv | undefined | null): boolean {
if (!env) return false;
return /^(1|true|yes|on)$/i.test((env.LOOPOVER_REVIEW_CONTENT_LANE ?? "").trim());
}
134 changes: 5 additions & 129 deletions src/review/content-lane/content-repo-spec.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,129 +1,5 @@
// Modular content-repository configuration for the curated-list content lane (the awesome-claude lane and any
// self-hosted curated list). The curated-list analogue of RegistryLaneSpec (the metagraphed registry lane): a
// maintainer whose list uses different categories or a different entry-file layout parameterizes the lane via
// config instead of a gittensory code change. Defaults preserve the awesome-claude behaviour byte-for-byte.
//
// This is a LEAF module (no content-lane imports) so every consumer — scope, duplicates, source-evidence — can
// import the spec without an import cycle. Fields are added here as each consumer is migrated.
export interface ContentRepoSpec {
/** The content categories the list accepts (the first path segment under the entry root). */
categories: ReadonlySet<string>;
/** Matches one content entry file, capturing [category, slug] — e.g. /^content\/([^/]+)\/([^/]+)\.mdx$/i. */
entryPathPattern: RegExp;
/** Head-branch prefixes used by bulk maintenance automation (link-health, etc.); these legitimately edit many
* entries in one PR and are ignored, never closed. */
maintenanceBranchPrefixes: readonly string[];
/** Frontmatter fields whose edit on a MODIFIED entry is a protected close — identity / provenance /
* verification / structural / monetization + supply-chain links. */
protectedFrontmatterFields: ReadonlySet<string>;
/** URL-bearing frontmatter keys (camelCase + snake_case) normalized + compared for duplicate detection. */
urlFields: ReadonlySet<string>;
/** Generic ecosystem hosts that never make a strict/aggressive domain-only match (a shared one is at most "related"). */
domainOnlyExclusions: ReadonlySet<string>;
/** Catalog roots that legitimately back MANY entries; a shared root alone is never strict (only a shared subpath). */
multiEntryCatalogUrls: ReadonlySet<string>;
/** Scalar source-URL frontmatter fields, in extraction ORDER (the source-evidence gate reads them in sequence). */
sourceUrlFields: readonly string[];
/** Array-valued source-URL frontmatter fields (e.g. retrievalSources/sourceUrls) read as real source evidence. */
sourceUrlListFields: ReadonlySet<string>;
/** Source fields treated as distribution (download/package) rather than canonical provenance. */
distributionSourceFields: ReadonlySet<string>;
/** Hosts that classify any source URL as distribution regardless of field (package registries / artifact hosts). */
distributionSourceHosts: ReadonlySet<string>;
/** Canonical fields that anchor the close decision + block inconclusive-downgrade (the primary provenance links). */
primaryCanonicalSourceFields: ReadonlySet<string>;
}

/** The default curated-list spec — awesome-claude's categories, entry layout, and maintenance branches. */
export const AWESOME_CLAUDE_CONTENT_SPEC: ContentRepoSpec = {
categories: new Set(["agents", "collections", "commands", "guides", "hooks", "mcp", "rules", "skills", "statuslines", "tools"]),
entryPathPattern: /^content\/([^/]+)\/([^/]+)\.mdx$/i,
maintenanceBranchPrefixes: ["links/"],
// PROTECTED = identity / provenance / verification / structural / monetization + supply-chain links. The entry's
// own REFERENCE/DOCS URLs are deliberately NOT protected (those links rot + legitimately need fixing); download/
// package/affiliate URLs stay protected (supply-chain / monetization risk).
protectedFrontmatterFields: new Set([
"affiliateUrl",
"author",
"authorProfileUrl",
"category",
"claimStatus",
"claimUrl",
"dateAdded",
"disclosure",
"downloadUrl",
"importPrNumber",
"importPrUrl",
"packageUrl",
"packageVerified",
"pricingModel",
"reviewedAt",
"reviewedBy",
"reviewedPrNumber",
"slug",
"submittedAt",
"submittedBy",
"submittedByUrl",
"sourceSubmissionNumber",
"sourceSubmissionUrl",
]),
urlFields: new Set([
"documentationUrl",
"docsUrl",
"downloadUrl",
"githubUrl",
"packageUrl",
"repoUrl",
"repositoryUrl",
"sourceUrl",
"websiteUrl",
"docs_url",
"download_url",
"github_url",
"package_url",
"repo_url",
"repository_url",
"source_url",
"website_url",
]),
domainOnlyExclusions: new Set(["github.com", "npmjs.com", "pypi.org", "raw.githubusercontent.com", "registry.npmjs.org"]),
multiEntryCatalogUrls: new Set([
"https://code.claude.com/docs/en/hooks",
"https://code.claude.com/docs/en/statusline",
"https://github.com/awslabs/mcp",
"https://github.com/microsoft/mcp",
"https://github.com/modelcontextprotocol/servers",
"https://github.com/snowflake-labs/mcp",
"https://github.com/twilio-labs/mcp",
]),
sourceUrlFields: [
"documentationUrl",
"docsUrl",
"downloadUrl",
"githubUrl",
"packageUrl",
"repoUrl",
"repositoryUrl",
"sourceUrl",
"websiteUrl",
],
sourceUrlListFields: new Set(["sourceUrls", "retrievalSources"]),
distributionSourceFields: new Set(["downloadUrl", "packageUrl"]),
distributionSourceHosts: new Set([
"crates.io",
"files.pythonhosted.org",
"hub.docker.com",
"marketplace.visualstudio.com",
"mvnrepository.com",
"npmjs.com",
"packagist.org",
"pkg.go.dev",
"plugins.gradle.org",
"pypi.org",
"registry.npmjs.org",
"repo1.maven.org",
"rubygems.org",
"www.npmjs.com",
]),
primaryCanonicalSourceFields: new Set(["githubUrl", "repoUrl", "repositoryUrl", "sourceUrl"]),
};
// content-repo-spec, extracted to @loopover/engine (#4880). Thin re-export shim; the implementation lives at
// packages/loopover-engine/src/review/content-lane/content-repo-spec.ts (imported via relative source path, not
// the published package, to match this repo's existing engine-consumption convention — see
// src/signals/check-summary.ts).
export * from "../../../packages/loopover-engine/src/review/content-lane/content-repo-spec";
28 changes: 4 additions & 24 deletions src/review/content-lane/flag.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,24 +1,4 @@
// Content-lane feature flag (convergence — reviewbot→gittensory content-review port).
//
// gittensory's native code-gate reviews CODE repos. The content lane reviews CONTENT repos — a
// curated list (awesome-claude) and a registry (metagraphed) — a different domain with its own
// deterministic primitives (duplicate detection, source-evidence reachability, security scanning,
// scope classification, and metagraphed's netuid grounding). The lane is ported as native,
// self-contained gittensory modules under this directory.
//
// FLAG-GATED + DEFAULT-OFF: the lane only runs when LOOPOVER_REVIEW_CONTENT_LANE is truthy in the Env.
// Flag-off, the host never reaches these modules, so the live behavior is byte-identical. At
// cutover the host flips the flag and routes awesome-claude + metagraphed PRs through the lane.

/** Env subset the content lane reads. The full Env adds it via env.d.ts; this keeps the lane
* testable without the whole binding (pass a plain object). */
export interface ContentLaneEnv {
/** When truthy ("1"/"true"/"on"/"yes"), the content lane is enabled. Default OFF. */
LOOPOVER_REVIEW_CONTENT_LANE?: string;
}

/** Is the content lane enabled? Default OFF — only a recognized truthy flag turns it on. */
export function isContentLaneEnabled(env: ContentLaneEnv | undefined | null): boolean {
if (!env) return false;
return /^(1|true|yes|on)$/i.test((env.LOOPOVER_REVIEW_CONTENT_LANE ?? "").trim());
}
// flag, extracted to @loopover/engine (#4880). Thin re-export shim; the implementation lives at
// packages/loopover-engine/src/review/content-lane/flag.ts (imported via relative source path, not the published
// package, to match this repo's existing engine-consumption convention — see src/signals/check-summary.ts).
export * from "../../../packages/loopover-engine/src/review/content-lane/flag";