stale-doc-counts-cause-conflict-cascade
When entity pages are added (e.g. alertmanager), the count ‘38 wiki pages’ in MEMORY.md, core-platforms.md, vault SKILL.md, and architecture.md all become stale simultaneously. Any single-file update that adds an entity, hook, cron, or MCP server must immediately grep all docs for the count reference and update every occurrence in the same operation. Stale counts are treated as conflicts, not cosmetic issues.
Related
- llm-agent-vs-cron-decision-hierarchy
- 2026-04-04-oracle-001-self-architecture-analysis
- enterprise-capability-expansion-5-pillars-from-digital-employee-analysis
- cache
- clawteam-openclaw-multi-agent-swarm-evaluation
- mcp-permanent-server-count-9-not-8
- orphaned-claude-sessions-ram-accumulation
- orphaned-claude-sessions-accumulate-600mb-ram-each
- stale-count-divergence-across-doc-files-pattern
- vault-wiki-entity-page-with-zero-source-count-is-valid
- memory-md-24kb-load-truncation-limit
- infra-service-count-stale-8-vs-10-actual
- memory-md-size-limit-causes-truncation-on-load
- infra-count-drift-opt-infra-vs-memory-md
- test-assertion-format-mismatch-markdown-vs-plain-text
- excluded-csvs-from-repo-bloat-and-language-reasons
- skill-count-cascade-both-architecture-md-and-memory-md
- brain-mcp-tools-fully-removed-use-vault-mcp