[Audit] [LLM] status file and README omit 8 existing Knowledge files; stale review metadata

[sergstack]

Finding (audit 2026-07-07)

ChatGPT/[LLM]/Knowledge/LLM_PROJECT_STATUS.md and ChatGPT/[LLM]/README.md both omit the same 8 Knowledge files that exist on disk:

• CONTEXT_ENGINEERING_PLAYBOOK.md
• CONTEXT_INTAKE_CHECKLIST.md
• CTC_PROMPT_STANDARD.md
• GOOD_BAD_CONTEXT_EXAMPLES.md
• LOCAL_AI_EXPERIMENT_PLAYBOOK.md
• LOCAL_AI_SECURITY_BOUNDARY.md
• LOCAL_MODEL_EVAL_MATRIX.md
• OLLAMA_OPENWEBUI_PILOT.md

Additional staleness in LLM_PROJECT_STATUS.md:

• last_reviewed: 2026-05-25 and current score: 8.6/10 were not refreshed even though the file itself was edited twice on 2026-07-07 (relationship + communication template batches).
• Acceptance checklist claims [x] README matches actual Knowledge files — currently false (see list above).
• Stated "Next fix: Keep README aligned with actual Knowledge files after future additions" was not applied during the latest additions.

Suggested fix

Docs-only PR: add the 8 files to the "Files present" list and to the README file inventory, refresh last_reviewed, and either update or un-check the acceptance line. Do not change any policy content.

Acceptance

• File lists in LLM_PROJECT_STATUS.md and README.md match ChatGPT/[LLM]/Knowledge/ on disk; metadata reflects the actual review date.