CB-622 Unit A: register fleet_* MCP tools, keep bridge_* working #135

Merged
ltms merged 1 commits from worker/cb-622a-165dff-1 into main 2026-08-22 21:58:03 +02:00
Member

Part of epic #125. Renames the documented MCP tool namespace from bridge_* to fleet_, while keeping every bridge_ name working for one release so a member spawned mid-migration does not break.

  • All eleven tools (ack, ask, list, poll, profiles, reply, send, spawn, status, stop, whoami) are now registered under fleet_*.
  • Each bridge_* name is registered as a twin against the SAME handler instance (no duplicated logic) with a description that starts with "DEPRECATED: use fleet_ instead."
  • A bridge_* call logs one WARN naming the old and new name, once per name for the life of the process (a Set, not a numeric sentinel).
  • REPLY_CHARTER in HerdrPeerLauncher.java now tells a spawned member to call fleet_reply (the rule that must reach a member with no repo checkout).
  • All other bridge_* string literals across mcp/, Javadoc comments, and tests were renamed to fleet_* for consistency.
  • No Java class/package/file was renamed; only string literals and new wiring/tests were added.

Build: mvn clean install from bridged/ -> BUILD SUCCESS, Tests run: 878, Failures: 0, Errors: 0, Skipped: 0 (up from the baseline 874; 4 new tests added in BridgeMcpTest for the dual registration + warn-once behavior).

Caveat: tests here are unit-level (proving registration + routing logic, and the warn-once mechanism via a captured logger). No live MCP client call was made from this worktree - that live verification is the lead's step per CB-618's lesson.

Part of epic #125. Renames the documented MCP tool namespace from bridge_* to fleet_*, while keeping every bridge_* name working for one release so a member spawned mid-migration does not break. - All eleven tools (ack, ask, list, poll, profiles, reply, send, spawn, status, stop, whoami) are now registered under fleet_*. - Each bridge_* name is registered as a twin against the SAME handler instance (no duplicated logic) with a description that starts with "DEPRECATED: use fleet_<x> instead." - A bridge_* call logs one WARN naming the old and new name, once per name for the life of the process (a Set, not a numeric sentinel). - REPLY_CHARTER in HerdrPeerLauncher.java now tells a spawned member to call fleet_reply (the rule that must reach a member with no repo checkout). - All other bridge_* string literals across mcp/, Javadoc comments, and tests were renamed to fleet_* for consistency. - No Java class/package/file was renamed; only string literals and new wiring/tests were added. Build: mvn clean install from bridged/ -> BUILD SUCCESS, Tests run: 878, Failures: 0, Errors: 0, Skipped: 0 (up from the baseline 874; 4 new tests added in BridgeMcpTest for the dual registration + warn-once behavior). Caveat: tests here are unit-level (proving registration + routing logic, and the warn-once mechanism via a captured logger). No live MCP client call was made from this worktree - that live verification is the lead's step per CB-618's lesson.
agent added 1 commit 2026-08-22 21:55:01 +02:00
CB-622 Unit A: register fleet_* MCP tools, keep bridge_* working
CI / build (pull_request) Successful in 1m2s
CI / contract (pull_request) Successful in 1m8s
41a3114d03
fleet_* is now the documented tool name for all eleven MCP tools; each
bridge_* twin is registered against the exact same handler (no logic
duplication) and its description leads with a DEPRECATED notice. A
bridge_* call logs one WARN naming the old and new name, once per name
for the life of the process (a Set, not a numeric sentinel).

REPLY_CHARTER in HerdrPeerLauncher now tells a spawned member to call
fleet_reply — the one rule that must survive with no repo checkout.

All other bridge_* string literals across mcp/, Javadoc, and tests were
renamed to fleet_* for consistency with the new documented name.
ltms merged commit 1deb0c90d4 into main 2026-08-22 21:58:03 +02:00
Sign in to join this conversation.