All three top-level counts are stale, tools by >3x. There's also an
internal inconsistency already present in the doc before this: the
Table of Contents says "Read Tools (20)" but the section header itself
says "### Read Tools (21)" — those never agreed with each other even
before comparing to the live server.
Given the doc has extensive per-tool documentation (categorized tables,
one row per tool/resource/prompt), a full accurate rewrite is a bigger
job than fits in a quick fix — filing this so the fuller pass is
tracked rather than silently living with 3x-stale counts. Whoever picks
this up should regenerate the counts and tables directly from
musehub/mcp/tools/__init__.py (MUSEHUB_TOOLS/MUSEHUB_TOOL_NAMES)
and the resources/prompts registries rather than hand-counting again.
Found while building Episode 16 ("Muse + MCP") demo material. Low severity — docs staleness, not a functional bug — but a large drift.
Live server (DEBUG=false, clean stdio transport):
docs/reference/mcp.md claims:
All three top-level counts are stale, tools by >3x. There's also an internal inconsistency already present in the doc before this: the Table of Contents says "Read Tools (20)" but the section header itself says "### Read Tools (21)" — those never agreed with each other even before comparing to the live server.
Given the doc has extensive per-tool documentation (categorized tables, one row per tool/resource/prompt), a full accurate rewrite is a bigger job than fits in a quick fix — filing this so the fuller pass is tracked rather than silently living with 3x-stale counts. Whoever picks this up should regenerate the counts and tables directly from
musehub/mcp/tools/__init__.py(MUSEHUB_TOOLS/MUSEHUB_TOOL_NAMES) and the resources/prompts registries rather than hand-counting again.Deferred per usual until after the season wraps.