Prompts

Everything the routine loads at runtime. The two master prompts (daily-cti-brief.md, weekly-summary.md) drive every run; the supporting files (verification.md, brief-template.md, check-brief-fixes.md) are read by the prompts at runtime as policy, template, and remediation guide.

Prompts and runtime policies

Version history

Every substantive prompt edit ships with a CHANGELOG entry explaining why the editorial policy shifted between two committed briefs. Recent versions:

2.43 — 2026-05-09 (per-agent model self-identification + Ops dashboard rework)
2.42 — 2026-05-09 (drop docs/improvements.md from the keep-current doc list)
2.41 — 2026-05-09 (separation of concerns: prompt-runtime files moved into prompts/, docs/ reworked, About page split into Documentation + Prompts)
2.40 — 2026-05-09 (docs cleanup: drop superseded historical/duplicate docs from prompt's keep-current list)
2.39 — 2026-05-09 (version-controlled auto-memory: .claude/memory/ + SessionStart symlink hook)
2.38 — 2026-05-09 (custom sub-agents: cti-research + cti-verification, model split Opus main / Sonnet workers, root CLAUDE.md)
2.37 — 2026-05-09 (rendering fixes: section numbering, Immediate Actions callout, UPDATE blockquote, footer parser, version banner)
2.36 — 2026-05-09 (publishing autonomy: feature-branch-only, auto-resolve conflicts, verify live site)
2.35 — 2026-05-09 (sync-then-publish chain — closes silent-skip on "main advanced")
2.34 — 2026-05-08 (weekly prompt brought up to daily quality bar — knowledge transfer)

Full version history →