agentleFS
lukesw55lukesw55/ai-pm-toolkit/.claude/skills/pm-transversal-comms/SKILL.md

pm-transversal-comms

Skill1 starsChanged 42 days agoMITView on GitHub

Cross-phase skill for **short-form PM communication** — executive email (SCQA), chat/Slack messages (BLUF), and channel-fit rules — the highest-frequency artefacts a PM writes. Invoke whenever a decision, status, or ask needs to go out as an email or chat message, when it's unclear whether something belongs in chat vs. email vs. a doc, when a thread is spiralling and needs to be pulled into a decision, or when an escalation needs a fact check before sending. Trigger on "write the email", "draft a Slack message", "quick update for the channel", "escalate this in chat", "should this be an email or a doc?", "e-mail para o time", "manda no Slack", "BLUF", "SCQA". Complements `pm-transversal-stakeholder` (longer-form memos/DACI) and `pm-storytelling` (narrative spine). Produces exec emails, chat updates/escalations, channel-fit recommendations. Chains into `humanizer` before anything ships.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When to use chat vs. email vs.

    line 55
  2. Very specific:

    Every phase: status updates, decision asks, and escalations happen constantly across Discover/Define/Develop/Deliver — this skill is the …

    line 78

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

PM Transversal — Short-form comms (email, chat, channel fit)

Transversal skill — useful in every Double Diamond phase whenever a PM needs to write, not present. Complements pm-transversal-stakeholder's longer-form memos/DACI and pm-storytelling's narrative spine with the two formats a PM actually reaches for most on a given day: email and chat.

Progressive loading

Load this SKILL.md first. For large or specialized tasks, use references/progressive-loading.md to choose the narrowest supporting reference before reading more.

When to use this skill

Invoke when:

  • a decision, status, or ask needs to leave as an email, not a doc
  • a chat message needs to escalate or land cleanly without turning into a meeting
  • it's unclear whether something belongs in chat, email, or a doc
  • an update needs to survive being read on a phone in ten seconds
  • a chat thread is spiralling and needs to be pulled into a decision or moved to a doc
  • an escalation is heated and needs a fact-check before it goes out

Skip for content that's long-form enough to need DACI/exec-memo structure (→ pm-transversal-stakeholder) or dedicated narrative work for a deck/QBR (→ pm-storytelling). Skip for Confluence/Jira artefacts (→ pm-transversal-docs).

Prime directive

Every message spends the reader's attention before it spends anything else. A well-structured 5-line Slack message that gets read beats a well-argued 20-line one that gets skimmed or ignored. Chat and email are not lightweight versions of a memo — they're a different contract with the reader, and that contract is: get to the point first.

Core sub-skills

1. Executive email (SCQA)

Situation–Complication–Question–Answer structure for decision, status, and ask emails to people with less time and context than the PM has.

Outputs: decision email, status email, ask email.

Anti-patterns: burying the ask below the fold, chronological narrative instead of SCQA, a subject line that doesn't name the action needed.

→ Deep-dive: references/exec-email-scqa.md

2. Chat messages (BLUF)

Bottom-line-up-front structure for Slack/Teams/WhatsApp: the point in line one, supporting detail threaded or collapsed, one ask per message.

Outputs: status update, escalation, quick decision ask.

Anti-patterns: wall of text with the point buried at the end, no explicit next action, letting a chat thread stand in for a decision record.

→ Deep-dive: references/chat-bluf.md

3. Channel-fit rules

When to use chat vs. email vs. a doc vs. a call; when and how to escalate a chat thread that's outgrown chat.

Outputs: channel recommendation, escalation trigger, hand-off to pm-transversal-stakeholder when the thread is actually a decision.

Anti-patterns: defaulting to whichever channel is already open, letting a negotiation sprawl across DMs with no written record, "let's hop on a call" as a way to avoid writing anything down.

→ Deep-dive: references/channel-rules.md

Workflow

  1. Load context — who's reading, what they already know, what decision or action this message needs to produce. When present, .ai/memory/org/company.md fixes the vocabulary and personas.md the reader's archetype.
  2. Classify the ask — email, chat, or "this actually needs a doc" (→ pm-transversal-stakeholder)?
  3. Name the one ask — one message, one action. Split multi-ask messages.
  4. Draft with the structure, not around it — SCQA for email, BLUF for chat. The structure is the discipline, not decoration.
  5. Check material claims — use ../inference-discipline/SKILL.md when facts are uncertain, changeable, variant-specific, or consequential. Reuse suitable evidence already available; recheck when freshness or scope requires it. Keep a material gap visible in the draft instead of turning it into certainty.
  6. Humanize before sending — route through ../humanizer/SKILL.md and the humanize-deliverables gate for anything leaving the workspace.

Output contract

Return the email or chat draft directly in the requested form. Add the reader, ask, channel, source, or open verification only when it changes the draft materially or the user requests that context. Do not force headings around a straightforward message.

Integration

  • Every phase: status updates, decision asks, and escalations happen constantly across Discover/Define/Develop/Deliver — this skill is the default for anything short-form.
  • Paired: pm-transversal-stakeholder (the thread/email escalates into a DACI or exec memo), pm-storytelling (a launch-comms email needs a narrative spine before SCQA structuring), inference-discipline (claims in a fast, emotional message are exactly where unverified assertions slip through), humanizer + humanize-deliverables (final pass before anything ships).
  • Doctrine: calibrated disagreement applies here too — a chat message under pressure to assert an unverified claim is the same failure mode as a solution-first premise in a discovery brief, just faster-moving. See ../DOCTRINE.md.

Communication modes follow ../../docs/patterns/COMMUNICATION_MODES.md. Per-skill: Lean (default) is BLUF/SCQA at the stated defaults; Standard adds the "why" a reader might ask for; Caveman is the ask and the date, nothing else.

Success criteria

  • the reader knows the ask and the deadline from the first line
  • nothing asserted as fact in a fast message is actually unverified
  • chat threads that are really decisions get pulled into a doc before they're 3 exchanges deep
  • emails get replies with decisions, not "can we hop on a call to discuss?"
  • escalations land as fact-checked and specific, not as blame with unconfirmed causes

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.