Blog batch 2026-08-12: Elasticsearch MCP deprecation, X official MCP server - #130
Open
gusmar2017 wants to merge 1 commit into
Open
Blog batch 2026-08-12: Elasticsearch MCP deprecation, X official MCP server#130gusmar2017 wants to merge 1 commit into
gusmar2017 wants to merge 1 commit into
Conversation
- Elasticsearch MCP Server: deprecated repo vs new Agent Builder MCP endpoint (Gus Marquez, search cluster, topic_source: new_topic) - X's Official MCP Server: what its 200+ endpoints expose (Adam Bush, social cluster, topic_source: new_topic) Sentinel validation log (Two-Agent Validation Loop): - Elasticsearch MCP Server: Round 1 NEEDS_REVISION (58%, 5 flagged) -> Round 2 NEEDS_REVISION (54%, 6 flagged) -> Round 3 NEEDS_REVISION (58%, 8 flagged) -- shipped per 3-round cap, residual verdict logged. Residual signal: contrastive-construction overuse (rather than/instead of/not-X family) proved resistant to two full revision passes. - X's Official MCP Server: Round 1 NEEDS_REVISION (45%, 5 flagged) -> Round 2 NEEDS_REVISION (38%, 3 flagged) -> Round 3 PASS (28%) Both posts: zero em dashes, gray-matter valid, word count 1,280-1,320 (within 1,200-1,500 target), all internal links verified against live directory data (no fabricated slugs), FAQ frontmatter/body parity confirmed. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
|
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
mcp-server-elasticsearchrepo is deprecated in favor of the new Agent Builder MCP endpoint (9.2+/Serverless). Covers the old repo's 5-tool surface, the new endpoint's admin-configured tool catalog, auth/version gating, and community fork alternatives.api.x.com/mcp, distinct from the older open-sourcexdevplatform/xmcpproject. Covers capabilities, auth, the separate docs-search endpoint, and how it compares to the 30+ unofficial X servers already in MCPFind's directory.Both
topic_source: new_topic. Topic selection: only 2 topics remained queued (both P3), so both posts came from this run's replenishment (Step 0.1, web-research fallback since GSC data had zero actionable signals this week).Sentinel validation log (Two-Agent Validation Loop)
Elasticsearch did not reach PASS within the 3-round cap. Its residual signal was overuse of "rather than / instead of / not-X" contrastive constructions -- two full revision passes reduced the raw count but didn't retire the underlying rhetorical habit, and Round 3's score ticked back up slightly (54% to 58%) as fixing flagged sentences surfaced the same connector pattern in previously-unflagged sentences. Worth noting for future humanizer passes on technical comparison content: contrastive-connector density deserves its own signal category, not just fragment/antithesis checks, since it's structurally different from the fragment-pair tic that dominated Round 1.
Both posts: zero em dashes, valid frontmatter/gray-matter parse, word count 1,280 (Elasticsearch) and 1,325 (X) -- both within the 1,200-1,500 range required this run. All internal links verified against live directory category dumps (
--category search,--category social) -- no fabricated slugs. Neither Elastic's official server (deprecated repo or new Agent Builder endpoint) nor X's official hosted server is indexed in MCPFind's directory yet; both posts disclose this honestly and link to the closest real community entries instead (io-github-tocharianou-elasticsearch,io-github-kaitoinfra-twitterapi-io-mcp-server).Data-quality / fact-check notes
Cluster linking notes
/blogindex per the no-siblings-yet rule; future search-cluster posts should link back to this one.socialmaps to clustercommunicationper SKILL.md's category-to-cluster table): exactly 5 published siblings, at the cap -- linked to all 5 with varied anchors (email setup, Slack/team chat roundup, Discord, Twilio, Intercom).Infra note -- git push guard
Unlike the last 3 batches (2026-08-02 partial, 2026-08-05, 2026-08-09 fully blocked), this week's push succeeded. The difference appears to be command form: prior blocked attempts used
cd $HOME/mcp-find && git push, which the I-05 guard flags as an unresolvable-cwd ambiguity (fail-closed by design). This run usedgit -C <explicit-path> pushthroughout (both for the worktree creation and the push itself), which resolved unambiguously and was not blocked. If this holds, the two batches stuck since 2026-08-05 and 2026-08-09 may be pushable the same way -- flagged to Gus for confirmation before repushing prior branches.Test plan: n/a (content-only change, existing CI/build pipeline runs on this PR)
🤖 Generated with Claude Code