hh-health-ai/healthcare-equity/comms-compliance

Client Communication & Compliance

Internal memos, IC briefings, client letters, translations, and the compliance guardrails around them. Prompts in ${CLAUDE_PLUGIN_ROOT}/skills/comms-compliance/references/prompts.md.

Route by output

(Quick-reaction earnings notes = COMM-03, in the earnings skill.)

Execution rules

1. Compression is the craft: COMM-01 caps at 1,000 words with the recommendation in three bullets up front; COMM-02 holds 600 words including "what we got wrong" (honesty is a feature, not a risk); COMM-04 requires an accurate analogy for the mechanism and two reasons a statistically positive result could be clinically irrelevant. Resist padding.
2. The compliance chain runs before anything leaves the desk: COMM-05 (flag statements derived from non-public sources; remove / rephrase-with-attribution / retain-with-compliance-review; err on caution) → COMM-06 (position disclosures long/short/none, personal holdings, meeting/call log, dated PTs, disclaimer) → COMM-07 when public-facing (restricted-window check, forward statements grounded in public data, investment-advice boundary). Run it whenever recent expert calls or management meetings touched the draft's subject.
3. Inputs come from the other skills — thesis memos, model outputs, post-print updates, EN-02's anonymised quotes — and every number keeps its time-stamp and source through the compression (standing instructions).
4. Structure of the IC memo mirrors the plugin's output skeleton (what matters now → facts → inference → model impact → variant view → disconfirming evidence); the "one question I want IC to answer" device from PORT-06 (thesis skill) belongs in every contested memo.
5. UK/EU stewardship-code and SFDR-adjacent reporting language pairs with the esg-stewardship skill (and the separate sfdr-si-prompt-library account skill if installed).