Agreement not yet measured: the pre-registered rule needs at least three agent pairs; this preview has fewer agents complete so far.
claude × codex: not computable (servers tied, no variance)
Pixel robot holding a trophy

DATABASES · CLAUDE CODE · SONNET 5.517 core tasks · claude-sonnet-5-5 · claude-code 2.1.296

BANDSERVERDETAILSCOMPLETION %
BAND 1 · statistically tied — listed A→Z
1DBHub (Bytebase)2 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
9482–100
1mcp-alchemy4 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1MCP Toolbox for Databases (Google)29 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1MotherDuck / DuckDB MCP4 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1mcp-server-mysql (benborla)1 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1mysql_mcp_server (designcomputer)3 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1postgresql-mcp-server (HenkDz)18 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1Postgres MCP Pro (crystaldba)9 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
9688–100
REFERENCE ROWS · archived/deprecated servers, scored but not ranked
REFserver-postgres (official, deprecated)1 tools · run log · repo coverage —% · reliability —% · extended 0%
median call 0.0s SHRED
100100–100
Completion = share of 17 core jobs finished, averaged per task over 2–5 attempts (95% interval below the score). Bands group servers whose rank intervals overlap. Coverage = share of capabilities the server advertises; reliability = pass rate on the jobs it claims to support.
Pixel robot holding a trophy

DATABASES · CODEX CLI · GPT-5.6 TERRA17 core tasks · gpt-5.6-terra · codex-cli 0.162.0 · LOW-SAMPLE

BANDSERVERDETAILSCOMPLETION %
BAND 1 · statistically tied — listed A→Z
1DBHub (Bytebase)2 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1mcp-alchemy4 tools · run log · repo · ⚑ >5% excluded attempts coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1MCP Toolbox for Databases (Google)29 tools · run log · repo · ⚑ >5% excluded attempts coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1MotherDuck / DuckDB MCP4 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1mcp-server-mysql (benborla)1 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1mysql_mcp_server (designcomputer)3 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1postgresql-mcp-server (HenkDz)18 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
1Postgres MCP Pro (crystaldba)9 tools · run log · repo coverage —% · reliability —% · extended 100%
median call 0.0s SHRED
100100–100
REFERENCE ROWS · archived/deprecated servers, scored but not ranked
REFserver-postgres (official, deprecated)1 tools · run log · repo coverage —% · reliability —% · extended 0%
median call 0.0s SHRED
100100–100
Completion = share of 17 core jobs finished, averaged per task over 2–5 attempts (95% interval below the score). Bands group servers whose rank intervals overlap. Coverage = share of capabilities the server advertises; reliability = pass rate on the jobs it claims to support.