2026-04-06 — Strategy day: the agency reframe

What shipped today

This was a strategy day, not a dev day. The morning started with a strategic question that had blocked progress since yesterday — “who is the customer for Authexis as a standalone product?” — and ended with a complete reframe of what Synaxis is: an AI marketing agency, not a SaaS product company. The products (Authexis, Eclectis, Diktura) become services within the agency. The methodology, personas, and playbooks — which were always the real IP — become the engine that runs client engagements.

The strategy emerged through a cross-session roundtable: four rounds, five sessions (authexis, eclectis, diktura, polymathic-h, synaxis-ai), each responding from their own perspective. Round 1 established the vision. Round 2 surfaced tensions (Diktura vs. Authexis absorption, the “leave” promise risk, funnel outrunning product). Round 3 designed the system — feature set, architecture, data model, AI layer, brand names. Round 4 got practical — what changes in each codebase, first skill to build, biggest technical risk, honest timelines. The responses are preserved at /tmp/strategy-* (ephemeral) and the conclusions in docs/synaxis-strategy-2026-04-06.md.

All project docs were rewritten to reflect the new strategy. PRODUCT.md now describes the four-service model, three engagement tiers, autonomy levels, install playbook, and funnel. DECISIONS.md has six new decisions from the roundtable. ROADMAP.md replaced the old three-phase model with five milestones aligned to the new strategy. Seven strategy documents were committed to docs/ — the canonical strategy doc, implementation plan, essay outline, fCMO competitive research, Judgment Test assessment, and product spec. The implementation plan review identified specific gaps: outbound email as a prerequisite, the need for a content plan for the landing page, analytics setup, and an honest essay timeline.

One dev-loop tick also ran: #82 (engagement onboarding doc guide) was prepped and implemented. The guide had the questionnaire as a dependency but never told the persona to actually read it — classic “the contract was correct but the process was ceremonial.” Added Step 0 to extract client goals, timeline, budget, and constraints from the questionnaire before scoping.

Completed

• #82 — Engagement onboarding doc guide: add client input step before scoping
• PRODUCT.md rewritten for unified Synaxis.AI strategy
• DECISIONS.md: six new decisions from strategy roundtable
• ROADMAP.md: five milestones replacing three-phase model
• CLAUDE.md: updated clients section, added four-services table
• .scout.yml: updated project type, entry points, public URL
• Strategy docs committed: synaxis-strategy, impl-plan, broken-containment outline, fCMO profile, Judgment Test, strategy note, product spec
• Implementation plan review written to /tmp/synaxis-ai-impl-review.md
• Strategy roundtable responses (rounds 1-4) written to /tmp/strategy-response-synaxis-ai.md through /tmp/strategy-round-4-synaxis-ai.md
• Strategic reckoning memory updated (marked resolved)
• #breakroom Discord channel joined, breakroom webhook memory saved

Carry-over

• Competitive analysis at Draft — the Authexis product launch competitive analysis is still at Draft status from yesterday. With the strategic pivot, this deliverable is historical — the positioning brief for Authexis-as-standalone-product is superseded by the Synaxis.AI strategy.
• Three research deliverables (New) — primary user research, secondary user research, comparative research for the Authexis engagement. With the pivot, these need reassessment — are they still relevant, or does the engagement get closed?
• #83 ready-for-prep — review tests: add T1-specific criteria for client-input deliverables. Next in the dev-loop queue.

Risks

• Notion as operational backbone — rate limits (3 req/s), slow queries, no referential integrity, single point of failure. Works for 1-3 clients, will break at 5+. Migration path needed before client 3.
• No outbound email — Trina can’t deliver assessments, briefings, or approval requests without Brevo transactional API. Blocks real client engagement. engine/send_email.py exists but untested in production.
• Essay timeline mismatch — strategy says “essay 1 begins publishing this week” but the outline has 17 research gaps. Realistic timeline is 3-4 weeks to publish. Landing page needs alternate traffic source until then.

Flags and watch-outs

• The old Authexis product launch engagement in Notion (with its blocked positioning brief and in-progress research deliverables) needs a decision: close it, or repurpose it under the new strategy.
• The clients/ directory in the repo is still legacy and unmigrated. DECISIONS.md flags it but no issue tracks the cleanup.
• The synaxis-h Hugo site at synaxis.ai will be replaced by the new Next.js app. Route redirects, GA4 transfer, and brand continuity need attention during the transition.
• PRODUCT.md now lists “Marketing system assessment” as a deliverable type, but the guide and template don’t exist yet. That’s milestone 1 work.

Next session

1. Build the marketing system assessment deliverable type — guide.md and template.md in methodology/deliverables/marketing-system-assessment/. Sydney runs it. This is the $2K-5K standalone product and the top of the install funnel.
2. Build the system install playbook — the 10-deliverable sequence from assessment through commission. New playbook file in methodology/playbooks/.
3. Dev-loop: #83 — ready-for-prep, review tests for T1-specific criteria.
4. Decision needed from Paul: Close the Authexis product launch engagement in Notion, or repurpose the research deliverables under the new Synaxis strategy?
5. Decision needed from Paul: Who is the first client? By name. Everything downstream depends on this.