# More Targeted Content Editing With Block MCP **Source doc**: [[3. Resources/Readwise/Podcasts/More Targeted Content Editing With Block MCP]] > [!question] Lingering Questions > - What exactly anchors a "stable reference ID" — block `clientId`, a persisted attribute, or a server-side map — and does it survive re-saves and revisions? > - Does the validation gate compare rendered HTML, serialized block markup, or both, and how does it behave with nested/innerBlocks, synced patterns, and third-party dynamic blocks? > - Is the reference-ID scheme interoperable with other MCP servers, or is it vendor-specific to GravityKit's implementation? > [!tip] Article Ideas > - A "how block-level MCP editing works" explainer (reference IDs, validation gate, revisions) for the Learn WordPress audience. > - A side-by-side demo comparing a generic "edit the HTML" MCP against a block-aware one with stable IDs + a render-validation gate. > [!link] Cross Reference > - Courtney's own `block-mcp-courtneyr-dev` project — worth auditing against this pattern: does it re-render through WordPress and save a revision per edit? > - [[AI Agent Safety MOC]] — render-validation-after-edit is a concrete instance of the general "don't trust the edit, verify it" pattern. > [!quote] Vocabulary > - **Stable reference ID** — an identifier Block MCP assigns to a block so a follow-up edit can target it directly without rescanning the page. > - **Update one block** — the specific MCP tool invoked in the demo; finds a block by reference ID and edits only that block. ## Author's thesis Zack Katz argues that block-level, ID-referenced editing plus a post-edit validation step that re-renders through WordPress itself is what separates disciplined AI content editing from generic "rescan and rewrite the HTML" MCP tools — the latter can silently corrupt block structure, while the former catches it and offers rollback via revisions. ## Key arguments (≤3) 1. **Stable reference IDs eliminate the rescan.** Other MCP products rescan the whole page after every edit to relocate what changed; Block MCP's stable IDs mean the tool already knows which block it's targeting, cutting time and token cost. 2. **Validation runs through WordPress's real render path**, not just HTML-syntax checking — confirming block structure survived the edit and throwing an error to the AI if it didn't. 3. **Every edit is a revision**, so a bad AI edit is one click from rollback — addressing the trust gap that makes site owners wary of write-scoped AI tools. ## Claim atoms ### Stable reference IDs remove the post-edit rescan > [!quote] ID:rw_block_mcp_1 > "We have stable reference IDs. That means that you don't have to rescan the page." Where/why: ~02:15, explaining what sets Block MCP apart from other MCP products. Matters because rescanning after every edit burns tokens and time; stable IDs are the efficiency mechanism behind "targeted" editing. ### Block MCP validates edits through WordPress's own render > [!quote] ID:rw_block_mcp_2 > "we run it through WordPress itself and make sure that WordPress renders the same way that we expect it to render." Where/why: ~03:20, the "verified" step after each edit. This is a render-truth check on block structure, not just HTML validity — the safety differentiator the whole demo is built around. ### Generic AI editing hunts for selectors and is slow > [!quote] ID:rw_block_mcp_3 > "AI can take forever to scan through and find some like specific code or look for a selector." Where/why: ~00:45, the problem statement Block MCP is built to solve — names the failure mode (selector/code hunting) that block-targeted editing replaces. ## Permanent candidates - **Validate AI content edits through the platform's real render path** — promoted below. The strongest claim: an editing tool earns trust by re-rendering its own edit and confirming structure, not by assuming success. - The "generic MCPs destroy block structure" comparison is a hasty generalization (no named competitor, no benchmark) — not promoted as a standalone claim, flagged instead as a fleeting fact-check. ## Processing checklist - [ ] Verify whether `block-mcp-courtneyr-dev` re-renders through WordPress and saves a revision per edit, following this pattern - [ ] Watch the full WP Minute Plus interview for the plugin supply-chain-security segment (referenced but not covered in this clip) - [x] Connect into the relevant MOC — [[3. Resources/MOCs/AI Agent Safety MOC]] ## Links - [[3. Resources/MOCs/AI Agent Safety MOC]] — render-validation-after-edit is a concrete, shippable instance of "don't trust the agent's self-report, verify the claimed state." - [[3. Resources/MOCs/AI & the Future of Knowledge Work MOC]] — targeted, block-aware editing is part of the shift from chat-mediated AI edits to editor-native AI actions.