agentsclimarketplace

Briefing source triangulation

Skill Ed3Design/ed3design-skill-bundles/planning-disciplines/skills/briefing-source-triangulation

Claude Code skill bundles for software engineering: 56 skills + 5 Python tools + 6 hooks + 4 sub-agents across 6 thematic plugins (token-savers, code-quality, planning-disciplines, async-forensik, schema-discipline, skill-system-meta). Empirically TDD-validated patterns, MIT licensed.

Install
npx -y skills add Ed3Design/ed3design-skill-bundles --skill briefing-source-triangulation

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when building a status briefing (morning standup, weekly review, "where was ", session-start briefing) that reads from multiple persistence sources (Daily Note + `.remember/today-*.md` + `.remember/recent.md` + git-log + file mtimes). Without conscious source-priority, briefings can claim items as "open" that were actually completed (e.g. email sent, meeting held, file shipped) because secondary sources lag or summarise. Trigger on phrases like "morning briefing", "daily standup", "where was ", "what was still open", "session-start briefing", "weekly review", "monthly close", "what's new since yesterday", "day overview", or any skill invocation that consumes `.remember/today-*.md` or `recent.md` as status input. Do NOT load when only consuming a single source (just reading the Daily Note for context), when writing INTO `. remember/` files (those are plugin-owned, not briefing input), or for code-only context-gathering (use git-log directly)

SKILL.md

4.4 KB, as published. Nobody here has run it

briefing-source-triangulation

The source hierarchy

When building any status briefing, sources rank in this priority order:

RankSourceWhat it isTrust for "did X happen?"
1Daily Note Daily Notes/YYYY-MM-DD.mdHuman/plugin-curated, intent-driven, structured blocksHigh for documented activity. Silent on external acts (sends, calls, meetings).
2Git log in code-reposAtomic commits with timestampsHigh for code changes. Silent on everything else.
3File mtimes on new artifactsFilesystem ground truthHigh for "when was this written". Silent on intent.
4User cross-check for external actsE-mail sent? Meeting held? Phone call done?Only the user knows. Ask, don't infer.
5.remember/today-*.md, recent.md, archive.mdPlugin-LLM-summarised secondaryPlausibility-check only. Do NOT treat as primary.

Why .remember/ is rank 5 (not rank 1)

The remember-plugin writes via post-tool/save-session hooks:

  • ## HH:MM | branch section markers are plugin-generated (likely session-start), not user-input times
  • Content is LLM-summarised (Haiku-cost signature visible in .remember/logs/memory-YYYY-MM-DD.log)
  • Sessions with <3 human messages are silently skipped (look for human msgs < 3, skip in the log)
  • Cross-day consolidation sessions can legitimately produce yesterday-themed entries under today's HH:MM marker (when the user explicitly asks for a "wrap up the open session" pass)

None of these are bugs. They are design choices that mean .remember/today-*.md cannot be the ground-truth source for "what happened today".

The triangulation procedure

Before writing any briefing:

  1. Read the Daily Note first. Skim blocks. This is the canonical chronicle.
  2. Check git log for repos relevant to the briefing scope: git log --since="midnight" --pretty=format:"%h %ai %s".
  3. For external acts the Daily Note mentions as "ready" / "ready to send" / "pending review" / "prepared" → ask the user whether they completed it. Status verbs like "ready to send" are claims of preparedness, not completion.
  4. Use .remember/today-*.md only as plausibility check. If something there contradicts the Daily Note, trust the Daily Note. If something appears under an HH:MM marker, don't assume the user typed at that time.
  5. Build the briefing. Carry-Over items inherit the source rank — if the only evidence for "X is open" comes from rank 5, mark it as "appears open, confirm with user".

The lesson (self-correction)

  • Daily Note said "ZIP ready to send" (rank 1, but a preparedness verb)
  • .remember/today-*.md had a stale block (rank 5)
  • No source said "email actually sent"
  • The briefing claimed "email send still open" → wrong, the email had been sent hours earlier
  • User had to correct it

The miss: the briefing should have asked the user "did the send happen?" instead of inferring "ready to send = still pending". The fix is operational: any Carry-Over claim about an external act gets a user-cross-check before it appears in the briefing.

Future refinements

These are deliberately non-blocking. The pattern works as documented; these would tighten it further:

  • A question template for external acts ("Did you send/meet/deploy X? When?")
  • A callable triangulation helper that emits a diff between Daily Note claims and .remember/today-*.md summaries
  • A briefing self-audit checklist (every claim has a source-rank annotation)

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.