docs: add security remediation plans from /improve audit - #323
Merged
Conversation
Adds 7 self-contained implementation plans (plus index) covering the highest-leverage security findings from a focused audit: unauthenticated MCP HTTP route, prototype-pollution writes in the data inspector, remote asset path traversal, OTP magic-link origin poisoning, symlink escapes in served/managed directories, and unpinned privileged GitHub Actions. No source code changed; plans/ is documentation for follow-up execution.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Adds seven self-contained implementation plans (plus an index) produced by a security-focused audit. No source code is changed — this is planning documentation for follow-up execution by implementer agents/humans.
Findings selected for plans, in recommended execution order:
plans/001-pin-github-actions.md— pin privileged GitHub Actions/reusable-workflow references to reviewed commit SHAs (release job currently grantscontents: write+id-token: writeto a mutable@mainreusable workflow).plans/002-authenticate-mcp-http.md— require an explicit bearer/callback authorization policy on the route-based MCP HTTP endpoint;Originis currently treated as identity, and@devframes/next/hubenables the route by default.plans/003-enforce-mcp-state-policy.md— apply the sameexposeSharedStatefilter to direct MCP resource reads, not just listing (depends on 002).plans/004-contain-remote-assets.md— validate remote-asset provider listings beforematerialize()writes them to disk, closing a path-traversal gap in build-time asset fetching.plans/005-block-data-inspector-prototype-writes.md— restrict Data Inspector's live-object write engine to own properties and reject prototype-sensitive keys.plans/006-validate-auth-link-origin.md— stop deriving the printed OTP magic-link origin from an unauthenticated first request'sHost/URL.plans/007-enforce-symlink-containment.md— canonicalize filesystem containment checks (static serving, assets plugin, open service) so pre-existing symlinks can't escape a configured root.Each plan is self-contained (current-state excerpts, exact scope, verification commands, test plan, done criteria, and STOP conditions) per the
improveskill's plan template, and was independently reviewed for executability before being finalized.Considered-and-rejected findings (dependency audit noise without a demonstrated reachable path, by-design OG/code-server behavior, token-lifecycle policy choices) are recorded in
plans/README.mdso they aren't re-audited.Test plan
git diff --checkpasses;pnpm lintran clean via the repo's pre-commit hook.This PR and the underlying plans were produced with the help of an AI coding agent (OpenCode).