Add Elasticsearch MCP server for PeaceLink site search #9
Labels
No labels
Doing
P1: critical
P2: important
P3: useful
P4: nice to have
P5: reminder
To Do
bug
invalid
wontfix
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
peacelink/ai-assistant#9
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Sibling repo, mirroring mcp-guardian. New ../mcp-elasticsearch/ FastMCP server exposing search over the existing ES cluster that indexes the full PeaceLink website. Register via MCP_SERVER_ELASTICSEARCH (or a dedicated MCP_ELASTICSEARCH_URL), add elasticsearch to the relevant research profiles' mcp_servers (start with RICERCA/CAMPAGNE, not all intents).
Open question to resolve during design: this ES index overlaps with the peacelink-* RAG sources already in pgvector (import_docs.py --source peacelink). Decide how to avoid the agent getting redundant hits for the same article from both paths — e.g. position ES as the fresher/complementary real-time source (like Guardian vs. static RAG) rather than a duplicate, or scope it to content/recency the RAG import doesn't cover.
Reference: RAG improvement discussion, [current conversation] — MCP recommended over the Wikipedia-style in-process tool because per-intent scoping and a potentially multi-operation surface (filter by category/date, fetch by ID) both fit the MCP pattern better.
Related to peacelink/pck3s#40