
from ai-first-sdlc-practices36
Generates a markdown dashboard of knowledge-base health: inventory, layer/domain distribution, recent activity, and staleness metrics.
Produces a read-only markdown report summarising the health of a knowledge base by reading shelf-index and log.md files. It computes inventory counts, layer and domain distributions, recent activity, and staleness indicators. Designed to be run locally in a repository that follows the SDLC knowledge-base layout.
Use this skill to audit a knowledge library before maintenance work, to produce periodic health reports, or as a preflight check before publishing or large refactors. It is read-only and does not invoke agent dispatch.
Best used with Python-capable tooling or local developer workflows; works well with code assistants that can run Python scripts and read repository metadata.
KB Stats generates a markdown dashboard of knowledge-base health metrics (inventory, layer/domain distribution, recent activity, staleness) by reading shelf-index and log.md files. No scripts were bundled for execution so analysis is static-only. The SKILL.md is well-structured with clear steps, argument hints, and explicit scope boundaries. Minor security note: the inline Python snippet uses dynamic module loading via importlib which carries mild injection risk if CLAUDE_PLUGIN_ROOT is untrusted, but this is standard plugin pattern. The skill is read-only and niche — only useful for teams already running the ai-first-sdlc-practices knowledge base plugin.
Clean, focused skill with no network calls, no destructive commands, no credentials. The inline python execution pattern is the only minor concern. Usefulness is limited to a specific ecosystem.