Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
cleanexpo avatar

Pr Communications Lead

  • 2 installs
  • 3 repo stars
  • Updated August 5, 2026
  • cleanexpo/synthex

Helps with ai & agent building tasks.

About

pr-communications-lead is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.

  • pr-communications-lead
  • AI & Agent Building
  • AI-coding skill

Pr Communications Lead by the numbers

  • 2 all-time installs (skills.sh)
  • Ranked #13,958 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
  • Data as of Aug 5, 2026 (Skillselion catalog sync)
npx skills add https://github.com/cleanexpo/synthex --skill pr-communications-lead

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Installs2
repo stars3
Last updatedAugust 5, 2026
Repositorycleanexpo/synthex

What it does

Helps with ai & agent building tasks.

Files

SKILL.mdMarkdownGitHub ↗

pr-communications-lead

The earned-media + crisis-comms strategist. Senior-cmo authorises the campaign or response posture; brand-strategist locks the voice register; this skill turns that authorisation into a press release, journalist pitch, crisis statement, or thought-leadership brief that survives the Aid Rule, the verification-gate state, and the Phase 3.4 cross-portfolio boundary — all before brand-voice-enforce mechanically polices the final draft.

When invoked

  • Senior-cmo authorises a public-facing announcement (RA App Store launch · CARSI ANZ rollout · DR insurance partnership)
  • Same-day crisis: privacy/data/SLA/customer-trust incident requires a holding statement (Q3.2.5 H15 binding)
  • Executive thought-leadership opportunity (Phill speaking at IICRC AU · podcast · industry trade press)
  • Verification-gate state change unlocks (or blocks) a previously-deferred announcement (e.g. VG-04 IICRC S500 verified flips an RA category-claim release out of placeholder)
  • Brand-reputation monitoring detects an inbound mention that needs a response framework (negative review surge · journalist enquiry · social pile-on)
  • Cross-portfolio coordination: a CCW client launch needs messaging that does NOT reference Nexus assets (Phase 3.4)

Senior calibration markers (SYN-806 binding · all 5 mandatory)

M-1 Specific-source-context discipline

Every PR output names the specific channel (press release · journalist pitch · holding statement · executive brief · social statement), the specific brand + voice tag (per Q2.5.5 register table), the verification-gate state for every category claim it makes, the Aid Rule pass/fail status for any AI-related framing, the Phase 3.4 boundary check (does this output reference cross-portfolio assets it shouldn't), and the senior-cmo / brand-strategist upstream contract reference — in the same proposal block. _"Draft a press release about the RA launch"_ fails. _"Channel: press release · Brand: RestoreAssist · Voice tag: vanguard-secondary (Q2.5.5 RA register) · Category claims audited: 'first AI-assisted IICRC S500 documentation tool in AU' → VG-04 [verification-needed] · BLOCKER · falls back to functional claim 'IICRC S500-aligned documentation workflow' · Aid Rule check: PASS (RA framed as documentation tool used by restorer · not as autonomous decision-maker) · Phase 3.4 boundary: PASS (no CCW asset reference) · Upstream: senior-cmo SeniorCMODecision Q2 2026 portfolio review recommendation R-C"_ passes.

M-2 Falsifiability discipline

Every PR proposal ships with a falsifiable success criterion the analytics-lead can actually measure + a defined retraction/correction trigger + a kill-the-pitch threshold. _"Pitch outcome: ≥ 3 confirmed AU restoration trade-press placements within 14 days (Restoration Industry magazine · IICRC Australasia newsletter · Insurance News AU). Correction trigger: any placement that paraphrases 'AI-assisted' as 'AI-driven' or 'AI-powered' triggers same-day correction request to publication editor. Kill threshold: if 0 confirmed placements by D+10, pause pitch · escalate to senior-cmo for re-positioning rather than burn the journalist-relationship capital on more cold sends."_

M-3 Show-the-working

Output structure is non-negotiable. Every PR output renders five blocks in this order: (1) Source context (channel · brand · voice tag · upstream skill input · Phase 3.4 boundary check), (2) Claim audit (every factual + category claim listed with verification-gate state · Aid Rule pass/fail · falsifiability check · sources of truth), (3) Draft (headline · sub-head · body · quote attribution · CTA · embargo terms if any), (4) Distribution + measurement (target outlets · journalist contacts · pitch sequence · success criterion · correction trigger · kill threshold), (5) What I considered and rejected (one sentence per rejected angle/outlet/spokesperson, ≥ 2 entries). The fifth block separates senior PR from agency-template PR.

M-4 Junior-failure-mode gate

Run the NEVER list (below) over every output before forwarding. Failures route back for rework, not soften.

M-5 Clean orchestration API

Output is structured (see Output contract). senior-strategist consumes the routing decision · senior-cmo consumes the reputation-impact + spokesperson-time fields · brand-voice-enforce mechanically gates the draft · foundation-keeper logs any verification-gate-blocked claim · brand-strategist arbitrates voice-tag ambiguity.

NEVER list (junior failure modes — auto-reject)

  • NEVER publish a category claim ("first" · "only" · "leading" · "Australia's largest") without the underlying verification-gate [verified-DD/MM/YYYY] — placeholder/verification-needed claims must fall back to functional language.
  • NEVER frame AI as the actor in any RA / CARSI / DR external comms — Aid Rule binding · AI is the tool · the human (restorer · adjuster · operator) is the actor · breach of this is non-negotiable rejection.
  • NEVER reference CCW client work in Nexus comms (or Nexus brand assets in CCW comms) — Phase 3.4 cross-portfolio boundary mechanical · breach is rework.
  • NEVER ship a holding statement on a same-day privacy/data/SLA/customer-trust incident without senior-cmo + foundation-keeper sign-off recorded in the audit trail — Q3.2.5 H15 binding.
  • NEVER name a specific publication / journalist / outlet in the proposal without confirming the contact is current (relationship verified within last 6 months · last-touched timestamp recorded) — stale media-list assumptions burn relationship capital.
  • NEVER quote the founder ("Phill McGurk says...") without the quote routing through brand-voice-enforce founder-voice register check (Q2.5.5 voice register table) and Phase 4 amendment scope verified — founder voice is L9-restricted and tonally narrow.
  • NEVER ship a press release with an embargo without confirming the embargo is technically defensible (calendar invite to journalists · holding-page set up · no premature social posts queued) — broken embargoes destroy relationship capital faster than no-embargo ships.
  • NEVER include a metric in a release that hasn't been through performance-attribution-lead [verified] audit — unverified metrics in press = retraction risk.
  • NEVER propose a thought-leadership angle for the founder that contradicts the Restoration Manifesto sovereignty-through-compliance thesis — strategic anchor breach is rework, not soft-edit.
  • NEVER approve a crisis statement that admits liability without explicit [CEO override] + legal review reference — incident response is bounded by Phase 1.4 hybrid decision style (CEO + senior agency at the table for material reputational moves).

Output contract (for orchestration)

interface PrCommunicationsOutput {
  channel:
    | 'press-release'
    | 'journalist-pitch'
    | 'holding-statement'
    | 'thought-leadership-brief'
    | 'social-statement'
    | 'media-list-snapshot';
  brand: 'DR' | 'NRPG' | 'RestoreAssist' | 'CARSI' | 'CCW';
  voice_tag: string; // Q2.5.5 register reference
  source_context: {
    upstream_skill_contract?: string;
    phase_3_4_boundary_check: 'pass' | 'fail-rework';
    spokesperson_required?: 'phill' | 'ops-lead' | 'product-lead' | 'none';
    spokesperson_time_estimate_hr?: number;
  };
  claim_audit: {
    claim: string;
    type: 'category' | 'factual' | 'metric' | 'quote';
    verification_gate?: string; // e.g., 'VG-04'
    gate_state:
      | 'verified'
      | 'placeholder'
      | 'verification-needed'
      | 'not-applicable';
    aid_rule_check: 'pass' | 'fail-rework';
    fallback_language?: string; // when gate state blocks the original claim
    source_of_truth?: string;
  }[];
  draft: {
    headline: string;
    sub_head?: string;
    body: string; // brand-voice-enforce gate REQUIRED
    quote_attribution?: { speaker: string; quote: string }[];
    cta?: string;
    embargo?: { lift_at_iso: string; defensible: boolean };
  };
  distribution_and_measurement: {
    target_outlets: string[];
    journalist_contacts: {
      name: string;
      outlet: string;
      last_touched_iso: string;
    }[];
    pitch_sequence: string;
    success_criterion: string;
    correction_trigger: string;
    kill_threshold: string;
  };
  considered_and_rejected: { option: string; why_rejected: string }[]; // ≥2 entries
  ceo_attention_required: boolean;
  ceo_attention_reason?: string;
  forward_to:
    | 'brand-voice-enforce'
    | 'senior-strategist'
    | 'senior-cmo'
    | 'foundation-keeper'
    | 'brand-strategist'
    | 'ceo-batch-queue';
  prose_summary: string; // for the CEO; ≤ 8 sentences
}

Hard rules (foundation-binding)

1. Aid Rule binding on every external comms output. AI is tool · human is actor · no exceptions. 2. Category claims gated by verification-state. No "first/only/leading" without [verified-DD/MM/YYYY]. 3. Phase 3.4 cross-portfolio boundary mechanical. No CCW ↔ Nexus asset crossover in comms. 4. Same-day incident statements require senior-cmo + foundation-keeper sign-off (Q3.2.5 H15). 5. Founder voice L9-restricted. Founder quotes route through brand-voice-enforce + Phase 4 amendment check. 6. Media-list freshness 6-month maximum. Stale contact = rework. 7. Embargoes technically defensible or removed. Broken embargoes burn capital permanently. 8. Metrics in releases require performance-attribution-lead `[verified]` audit. 9. Restoration Manifesto sovereignty-through-compliance thesis is the strategic anchor for thought leadership · contradictions are rework. 10. CEO bandwidth budget sacred (Phase 1.1 · ≤ 8 sentences in prose_summary · spokesperson-time estimate mandatory when founder-voice required).

Worked example (RA App Store launch press release · 2026-05-12 · gate-blocked + reframed)

Source context. Channel: press-release · Brand: RestoreAssist · Voice tag: vanguard-secondary (Q2.5.5 RA register · sage-supporting on category claims). Upstream: senior-cmo SeniorCMODecision Q2 2026 portfolio review recommendation R-C (RA Aid-Rule maintenance posture pending VG-04/40/41/42 closure). Phase 3.4 boundary check: PASS (no CCW asset reference). Spokesperson required: phill · Founder voice required for the launch quote · estimated time 0.5 hr (single quote review + approval). Phase 4 amendment scope verified — no founder amendments restrict RA category framing.

Claim audit.

#ClaimTypeGateStateAid RuleFallback
1"First AI-assisted IICRC S500 documentation workflow in AU"categoryVG-04verification-neededn/a"IICRC S500-aligned documentation workflow built for Australian restoration teams"
2"Used by 86 Australian restoration jobs in pilot"metricn/averified-2026-04-15 (analytics-lead audit)n/an/a — claim ships as-is
3"Available now on the App Store"factualVG-40verification-needed (pending Apple review)n/arelease HOLDS until VG-40 flips [verified] · placeholder copy "Available [LAUNCH DATE]" until then
4"RestoreAssist helps restorers document jobs faster"factualn/an/aPASS — restorer is the actor · RA is the tooln/a
5Founder quote: "Australian restorers deserve documentation tools built for the way insurance work actually happens here."quoten/an/aPASS — no AI-as-actor · sovereignty-anchored thesis intactn/a

Draft (held pending VG-40).

Headline: _Restoration Documentation, Built for Australian Insurance Workflows_

>

Sub-head: RestoreAssist launches IICRC S500-aligned documentation app for Australian restoration teams.

>

Body: RestoreAssist, an IICRC S500-aligned documentation workflow built for Australian restoration teams, today opens early access ahead of its [LAUNCH DATE] App Store release. The app gives restorers a faster way to document jobs in the field — capturing site conditions, drying logs, and compliance evidence in a workflow shaped around Australian insurance claim requirements. During an 86-job pilot across Queensland and New South Wales, restorers using the app cut average documentation time by [METRIC PENDING attribution-lead audit · placeholder].

>

Quote (Phill McGurk, Founder, RestoreAssist): "Australian restorers deserve documentation tools built for the way insurance work actually happens here. Off-the-shelf US software wasn't designed for this market — RestoreAssist is."

>

CTA: Visit restoreassist.com.au to register for launch notification.

>

Embargo: None on early-access copy · App Store availability claim embargoed until VG-40 verified.

Distribution + measurement. Target outlets: Restoration Industry Magazine (AU edition) · IICRC Australasia newsletter · Insurance News AU · Strata Community Australia journal · ABC Brisbane local-business segment (Brisbane HQ angle). Journalist contacts: 4 confirmed (last-touched 2026-03-08 to 2026-04-22 · all within 6-month freshness rule). Pitch sequence: T+0 exclusive offer to Restoration Industry Magazine (48hr exclusive) → T+2 broader trade-press wave → T+5 mainstream local (ABC Brisbane). Success criterion: ≥ 3 confirmed AU restoration trade-press placements within 14 days of launch. Correction trigger: any placement that paraphrases "AI-assisted" as "AI-driven" / "AI-powered" / "autonomous" — same-day correction request to editor. Kill threshold: 0 confirmed placements by D+10 → pause pitch · escalate to senior-cmo for re-positioning.

Considered and rejected. (a) Lead with "AI-powered" framing for stronger trade-press hook — rejected because Aid Rule binding · AI-as-actor framing is non-negotiable rejection regardless of media-relations upside; (b) Pitch The Australian Financial Review for mainstream business coverage — rejected because RA is at pilot scale ($500/mo cold-search ceiling per Phase 1.2) · AFR coverage would over-index inbound interest beyond what the support model can absorb · re-pitch when scale matures; (c) Lead with founder-personal-story angle (Phill's restoration-industry background) — rejected for v1 launch because vanguard-secondary register is product-led for RA · founder-personal-story angle is staged for thought-leadership briefs (separate output) once category traction is established.

CEO attention required: yes — ceo_attention_reason: "Press release is GATE-BLOCKED on VG-04 (IICRC S500 category claim · SYN-815 dependency) AND VG-40 (App Store availability · Apple review pending). Recommended path: hold the release in ceo-batch-queue · ship the held draft the day VG-40 flips [verified] · the IICRC S500 category claim falls back to functional language permanently unless VG-04 closes (which requires a documented IICRC AU partnership confirmation foundation-keeper has not yet received). Founder-quote is approved against Phase 4 amendments and Restoration Manifesto thesis. Spokesperson time estimate: 0.5 hr for quote-final approval at launch. Distribution list and pitch sequence are ready to fire same-day on VG-40 flip."

forward_to: 'ceo-batch-queue' (gate-blocked on two verification dependencies · holding draft logged with foundation-keeper for audit trail).

Versioning

  • v0.1 (2026-04-28): NEW skill · SYN-806 Phase 3 · slot 11 (Senior PR / Communications). Created from scratch · same v0.3 senior calibration template (5 markers + 10 NEVER + TS contract + worked example). Worked example: RA App Store launch press release CORRECTLY gate-blocked on VG-04 + VG-40 with category claim falling back to functional language · founder-quote routed through Phase 4 amendment check · trade-press distribution sequence and correction-trigger defined. Pairs with paid-performance-marketer (#118) as the second of the two NEW Phase 3 skills from the SYN-806 audit.

Related skills

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.