solanasis-content-creation-posting-strategy-research-playbook-handoff-2026-03-18

Executive Summary

This document extracts, verifies, and improves the key points from the discussion about how Solanasis should approach content creation and social posting in 2026, with an emphasis on LinkedIn first, then X, Instagram, and one additional compounding channel.

The strongest evidence-backed conclusion is that Solanasis should not try to build a brittle, fully autonomous “AI posts everywhere” machine. The current, credible playbook is a human-in-the-loop, AI-accelerated content system: use AI and browser agents for research, idea extraction, repurposing, drafting, metadata preparation, workflow automation, and queue management; use official APIs or a scheduler for publishing; and keep a human responsible for final approval, comments, DMs, and judgment.

The most defensible operating model for Solanasis today is:

Primary channel: LinkedIn, led from Dmitri’s personal profile, supported by the Solanasis company page.
Secondary channel: X for sharper operator takes, news reactions, and network-building.
Support channel: Instagram for credibility, culture, founder clips, and visual proof-of-work rather than primary demand generation.
Additional channel: YouTube as the highest-compounding long-form library and repurposing source.
Operational stack: Claude Code for research and drafting; a scheduler/API layer such as Postiz (strongest AI-native fit), with Buffer / Planable / Metricool as alternatives depending on workflow priorities.
Workflow principle: AI upstream, API/scheduler at publish time, human in the loop for approvals and engagement.

This document also adds missing but important considerations that were not fully addressed in the original discussion, including: platform-specific API constraints, approval workflows, voice preservation, analytics feedback loops, token security, channel-role clarity, and what another AI or contractor would need in order to continue this work without rereading the original chat.

Purpose of This Document

This artifact is meant to serve simultaneously as a:

Guide for Solanasis social/content strategy
Playbook for execution
Briefing memo for a future AI or contractor
Handoff document that preserves verified facts, constraints, and open questions

This is not just a narrative recap. It is a structured operational reference.

Evidence Status Legend

Every important point in this document is labeled as one of the following:

Verified — confirmed from a cited reliable source, usually official documentation or a high-confidence primary source
User-stated — comes from the user or user-provided materials and should be treated as context rather than externally verified fact
Assistant-stated but unverified — appeared in the earlier discussion but could not be fully verified from reliable sources
Tentative / speculative — recommendation, inference, hypothesis, or practitioner belief that may be useful but is not a hard fact

Discussion Context

User goals and constraints

User-stated: Solanasis wants a proper content creation and posting strategy focused mainly on LinkedIn, with some use of Instagram and X, and possibly one additional channel more likely to attract clients.
User-stated: The user wants to understand what people are actually doing in 2026, especially what marketing agencies are running for clients.
User-stated: The user wants an AI-native agency model, including what can be done through Claude Code and Claude Code’s Chrome/browser capabilities from a technical standpoint.
User-stated: The user cares about API-accessible workflows and wants to know what builders on Reddit and related user-reported communities are actually using.
User-stated: The user wants something operationally useful, not generic theory.

Brand / voice / operating context preserved for future use

User-stated: Solanasis is positioned around security, resilience, and operations.
User-stated: Solanasis should sound credible, founder-led, clear, practical, and not like generic AI marketing.
User-stated (from uploaded style reference): Dmitri prefers writing that is succinct, usually no more than 1–3 sentences per paragraph, avoids sensational “this isn’t X, it’s Y” framing, and favors grounded, practical prose over bloated copy.
User-stated (from uploaded style reference): The “later” Dmitri voice is warmer, more practical, more business-grounded, and more results-oriented than earlier, more spiritual writing.
User-stated (inferred from prior project context): Solanasis is a founder-led B2B services firm targeting SMBs, nonprofits, and adjacent trust-based buyers who need operational resilience, security, systems cleanup, AI enablement, and executive-level operational partnership.

Key Facts and Verified Findings

1) Claude Code can assist with browser-based content operations, but it is not the ideal final publishing backbone

Verified: Anthropic documents that Claude Code can connect to Chrome to automate form filling, extract data from web pages, draft content in Google Docs, and run multi-site workflows.[1]
Verified: Anthropic also states that Claude Code browser actions run in a visible Chrome window and share the browser’s existing login state.[1]
Verified: Anthropic’s help center says Claude in Chrome is in beta and available on paid plans, and that it can read, click, and navigate websites alongside the user.[2]
Verified: Anthropic explicitly warns that browser use carries inherent risks and should be used carefully.[2]
Tentative / speculative: This makes Claude Code strong for supervised content operations — research, drafting, uploading assets, populating tools, quality checks, and light workflow automation — but weaker as the sole unattended “publish everything everywhere” layer.

Why it matters: The earlier discussion correctly pointed toward using Claude Code upstream of publishing. That direction holds up technically. It is a useful assistant, not the most stable system of record for unattended social publishing.

2) Official platform APIs exist for core publishing use cases, but their constraints differ by platform

Verified: LinkedIn’s Posts API supports posting on behalf of an organization with w_organization_social if the authenticated user has the appropriate company-page role.[3]
Verified: LinkedIn’s Posts API also supports posting on behalf of an authenticated member with w_member_social.[3]
Verified: Some LinkedIn read permissions are restricted or approval-gated, including r_member_social.[3]
Verified: LinkedIn’s business resources continue to position LinkedIn Pages as a way to establish thought leadership, build brand awareness, and move audiences to take action.[4]
Tentative / speculative: The practical implication is that LinkedIn supports both personal-profile and company-page posting, but its permission model and partner restrictions make direct integration work less straightforward than some other platforms.

X

Verified: X’s current API includes a “Create or Edit Post” endpoint for programmatic posting.[5]
Verified: X’s API documentation states that the API supports publishing content and uses modern REST endpoints with pay-per-usage pricing.[6]

Instagram

Verified: Meta documents official Instagram content publishing support through the Instagram platform / Graph API for single images, videos, reels, and carousel posts.[7]
Tentative / speculative: Instagram is automatable through the official API, but implementation details and account prerequisites typically make it more cumbersome than users expect. It is better treated as a supported publish target than as the center of the system.

YouTube

Verified: Google documents uploading videos through the YouTube Data API.[8]
Verified: Google’s YouTube API reference indicates that unverified API projects that upload videos will have them set to private by default until they pass audit.[9]

Why it matters: The publish layer should be built around platform-supported APIs or mature schedulers that wrap them, rather than around fragile UI automation alone.

3) The strongest “AI-native” publishing/scheduling fit discussed was Postiz, but it should still be treated as a tool choice, not a settled fact

Verified: Postiz documents a CLI for automating social media posting from the command line and says it wraps the Public API.[10]
Verified: Postiz’s official site and GitHub materials position it as a social-media scheduling tool with a public API, AI-agent orientation, and support for many platforms.[11][12]
Verified: Postiz’s GitHub README states that its hosted service uses official platform-approved OAuth flows and does not automate or scrape content from social media platforms.[12]
Verified: Postiz supports both cloud and self-hosted deployment paths and offers public API authentication via API key or OAuth2 token.[13]
Verified: Postiz’s public API documentation lists support for numerous platforms including X, LinkedIn, LinkedIn Page, Instagram, Reddit, YouTube, TikTok, and others.[13]
Verified: Postiz’s public API docs currently state a 30 requests per hour limit on all endpoints.[13]
Tentative / speculative: Postiz is a strong fit for an AI-native Solanasis workflow because it is API-first, agent-friendly, and self-hostable. However, it still needs real validation against Solanasis needs, especially around reliability, approval UX, analytics depth, and LinkedIn specifics.

Why it matters: Postiz is not “proven best” in an absolute sense, but it is the strongest technically coherent recommendation from the earlier discussion.

4) Buffer, Planable, and Metricool remain relevant, but for different reasons

Buffer

Verified: Buffer’s legacy developer docs still expose an API for pending/sent updates, profiles, and scheduled times, with published rate limits.[14]
Verified: Buffer also states that it is rebuilding its public API and that the new developer API is in beta, accessible through Buffer settings.[15]
Tentative / speculative: Buffer looks viable if Solanasis wants something established and simpler, but its API transition status is a meaningful caveat.

Planable

Verified: Planable explicitly positions itself around creating, reviewing, approving, and scheduling social content in one place, with in-context previews and approval workflows.[16]
Tentative / speculative: Planable is stronger for review and collaboration than for “AI-native CLI-first automation.”

Metricool

Verified: Metricool has an API/integrations help center and markets itself as a scheduling/analysis platform.[17][18]
Assistant-stated but unverified: The earlier discussion implied Metricool’s API access depends on higher-tier plans; this may be true, but it was not fully verified from a primary source during this pass.
Tentative / speculative: Metricool may be acceptable for smaller operations and broad publishing/analytics, but agency-level team workflow fit should be tested directly.

Why it matters: These are not interchangeable.

Choose Postiz for API-first / AI-native / self-hosted flexibility.
Choose Planable for approval-heavy client workflows.
Choose Buffer for a more established but currently transitional publish API experience.
Choose Metricool for a broader all-in-one scheduler/analytics setup if the workflow is simple enough.

5) The most defensible channel hierarchy for Solanasis is LinkedIn first, X second, Instagram support, YouTube compounding

LinkedIn first

Verified: LinkedIn officially encourages Pages as thought-leadership assets for businesses.[4]
Assistant-stated but unverified: The earlier answer claimed that “personal profile first, company page second” is the practical B2B pattern.
Tentative / speculative, supported by practitioner reports: Multiple recent Reddit threads from marketers and founders say personal profiles feel more alive and effective than faceless company pages, while company pages still matter as a legitimacy layer and archive.[19][20][21]

X second

Verified: X supports programmatic posting and remains technically open enough for AI-assisted workflows.[5][6]
Tentative / speculative: For Solanasis specifically, X is likely useful for sharper operator takes, ecosystem networking, event/news reactions, and visibility among founders and tech-adjacent operators.

Instagram as support

Verified: Instagram content publishing is officially supported for media posts.[7]
Tentative / speculative: For a trust-based B2B services firm, Instagram is more plausibly a credibility and culture channel than a primary direct-response lead-gen engine.

YouTube as the additional “compounding” channel

Verified: YouTube’s official API supports video uploads and management.[8][9]
Tentative / speculative: For Solanasis, YouTube is the strongest additional channel because one talking-head video or walkthrough can be repurposed into LinkedIn posts, X threads, clips, carousels, newsletters, and website content.

Why it matters: This hierarchy is not just about where to post. It dictates what kind of content to produce and what infrastructure to build.

6) Agency and practitioner workflow evidence supports a smaller core stack, approvals, and human review — not maximal tool sprawl

Verified (official vendor positioning): Planable emphasizes review, approval, and scheduling as a core workflow.[16]
Tentative / speculative, supported by user-reported evidence: Recent Reddit discussions in social-media and marketing communities consistently recommend using one solid scheduler, basic analytics, and a simple collaboration layer rather than an overgrown stack.[22]
Tentative / speculative, supported by user-reported evidence: Recent Reddit threads also describe workflows where Claude/Claude Code handles drafting and operational admin, while scheduling and cross-posting run through dedicated tools.[23][24][25]

Why it matters: Solanasis should resist the urge to overbuild. The winning system is usually simpler than founders expect.

Major Decisions and Conclusions

A. Recommended strategic direction

Tentative / speculative conclusion: Solanasis should adopt a founder-led authority model rather than a “brand account only” model.
Tentative / speculative conclusion: LinkedIn should be the primary organic channel, with Dmitri’s personal profile as the lead account and the Solanasis page as supporting infrastructure.
Tentative / speculative conclusion: Solanasis should not try to make Instagram carry the business. Use it as a lightweight trust and proof channel.
Tentative / speculative conclusion: Solanasis should add YouTube before adding more text-based channels.
Tentative / speculative conclusion: Reddit should be used primarily for listening, research, topic mining, and selective helpful participation, not as the main publishing channel.

B. Recommended technical direction

Tentative / speculative conclusion: The best current technical model is:
- Claude Code for research, synthesis, drafting, repurposing, and workflow glue
- Postiz as the first tool to test for scheduling and publishing
- Official platform APIs or a mature scheduler as the publish layer
- Human approval before posting
- Manual human engagement after posting

C. Recommended operating principle

Tentative / speculative conclusion: The strongest principle from the discussion is:

AI should do the heavy upstream work; humans should own final judgment, relationship-building, and learning loops.

Reasoning, Tradeoffs, and Why It Matters

1) Why founder-led beats “faceless brand” for Solanasis

Tentative / speculative: Solanasis sells trust-heavy services: security, resilience, operational cleanup, AI enablement, and executive-level partnership.
Tentative / speculative: These are bought from people who need confidence in judgment, not just awareness of a logo.
Tentative / speculative: A founder-led profile can carry specificity, opinions, field observations, and story-driven credibility better than a company page alone.

Tradeoff:
Founder-led content is harder to fake and more powerful when done well, but it requires actual consistency and thought. It also creates dependence on Dmitri’s time and voice. That risk must be managed through systems, prompts, and repurposing workflows.

2) Why AI should sit upstream of publishing

Verified: Claude Code can automate browser tasks, drafting, and data extraction.[1][2]
Verified: Platform APIs exist for formal publishing.[3][5][7][8]
Tentative / speculative: Browser automation alone is less stable than platform APIs or a scheduler because page layouts change, sessions expire, and CAPTCHAs / permissions interrupt flows.
Tentative / speculative: Using AI for ideation, repurposing, structuring, metadata, and queue preparation captures most of the leverage without making the system unnecessarily brittle.

Tradeoff:
A fully autonomous system feels exciting, but a supervised system will usually outperform it operationally for a small, trust-based services firm.

3) Why one extra channel should be YouTube, not “everything else”

Verified: YouTube has strong official upload support.[8][9]
Tentative / speculative: A single 6–12 minute YouTube video can be atomized into many other assets.
Tentative / speculative: That makes YouTube the best “source-content engine” for a lean AI-native team.

Tradeoff:
YouTube is more production-heavy than text posting, but if Solanasis can create even 2 useful videos per month, the downstream repurposing payoff is substantial.

4) Why approval workflows matter more than “best AI model”

Verified: Planable’s positioning centers on approvals, previews, and scheduling.[16]
Tentative / speculative, practitioner-backed: User-reported workflows repeatedly point to approval friction as the place where content systems break, not merely draft quality.[24][25]
Tentative / speculative: For Solanasis, content velocity will depend less on generating ideas and more on making approval, revision, and publication friction low enough to sustain weekly output.

Recommended Playbook / Process

1. Operating model

Core principle

Use a source-content → AI repurposing → approval → publish → engagement → analytics feedback loop.

Roles

Dmitri / founder: source ideas, approve final posts, record videos, engage in comments/DMs
AI system (Claude Code + prompts + supporting tools): research, repurpose, draft, summarize, format, prep metadata, organize queues
Scheduler / API layer: publish reliably and track status
Optional human assistant / contractor: calendar upkeep, asset cleanup, clipping, reporting, approvals routing

2. Channel strategy

Role: primary pipeline and authority channel

Recommended use

Founder POV posts
Practical breakdowns from the field
Security / resilience takes
“What actually breaks” posts
Short contrarian-but-grounded observations
Light document/carousel posts where useful
Event notes, client-pattern insights, and lessons learned

Cadence

Tentative / speculative recommendation: 3 high-quality founder posts per week
2 lighter posts or reposts from company / team / events as capacity allows
Daily or near-daily commenting on relevant posts

Important note

Tentative / speculative: The company page should exist, stay active, and provide legitimacy, but it should not be treated as the primary discovery engine.

X

Role: network density, fast takes, ecosystem visibility

Recommended use

1–2 original posts per day
2–5 thoughtful replies per day
Thread versions of strong LinkedIn insights
News reactions about AI, security, operations, resilience, nonprofit tech, SMB chaos, and responsible AI

Caution

Tentative / speculative: X can consume attention fast. Treat it as a high-signal side channel, not the center of the strategy.

Instagram

Role: visual trust, culture, founder presence, behind-the-scenes proof

Recommended use

Founder clips
Carousels from LinkedIn posts
Short reels from videos
Event / meeting / workshop / whiteboard / setup visuals
Brand presence, not heavy sales copy

Caution

Tentative / speculative: Keep Instagram intentionally lightweight unless evidence later shows it is producing real conversations or referrals.

YouTube

Role: compounding source-content engine

Recommended use

2–4 videos/month
Topics tied to operational resilience, security reality, AI enablement, systems cleanup, founder/operator lessons
Talking-head, walkthrough, whiteboard, screen-share, or case-study style
Each video becomes multiple downstream assets

Role: listening, research, and selective helpfulness

Recommended use

Mine recurring pain points in SMB ops, nonprofit tech, compliance, cybersecurity, AI implementation, CRM/system cleanup, and founder operations
Capture language people use
Convert pain patterns into platform-specific content
Participate only where you can genuinely help

3. Content pillars for Solanasis

Below are the strongest candidate pillars extracted and improved from the discussion.

Pillar 1 — Operational resilience reality

What actually breaks in SMBs and nonprofits
Why “backup exists” is not the same as “recovery works”
Fragile systems hiding behind okay-looking dashboards
Lessons from real operational messes

Pillar 2 — Security without theater

Security basics that are still being skipped
False confidence created by checkbox security
What leaders think they bought vs what they actually have
Calm, grounded explanations of risk

Pillar 3 — Messy systems into clean operations

CRM cleanup, integrations, data migrations, permissions chaos, process drift
Before/after thinking
Architecture that respects lean teams

Pillar 4 — Responsible AI for real operators

Where AI is worth it
Where AI introduces risk or operational debt
How to stay human-in-the-loop
Practical automation, not hype

Pillar 5 — Founder/operator perspective

What Solanasis is seeing in the market
What smart SMB and nonprofit operators should stop tolerating
Field notes, mistakes, and sharp observations
Building a serious firm with lean systems and AI leverage

4. Content formats

Story + lesson post
“What we keep seeing” pattern post
Myth / misconception clarification
Practical checklist post
“Behind the scenes of a systems cleanup” post
Founder memo post
PDF/carousel for frameworks or before/after process maps

X

Sharp observation
Contrarian-but-grounded take
Thread summary of LinkedIn post
Reply chain around current tech/ops/security issues

Instagram

Carousel distilled from LinkedIn
Reel clipped from YouTube / Loom / Zoom
Founder photo + caption
Event or workshop recap

YouTube

Talking-head lesson
Screen-share walkthrough
Whiteboard breakdown
Case-pattern analysis (with confidentiality preserved)

5. Weekly workflow

Monday — Source capture and prioritization

Pull ideas from:
- client work patterns
- discovery calls
- voice notes
- internal Slack / email / docs
- Reddit pain points
- LinkedIn comments and DMs
- industry news
Rank ideas by:
- relevance to target buyers
- specificity
- credibility
- ease of production
- repurposability

Tuesday — Draft and repurpose

Claude Code takes 3–5 source ideas and turns them into:
- 2–3 LinkedIn drafts
- 5–10 X drafts or thread candidates
- 1 Instagram carousel outline or caption
- 1 short video outline or clip plan
Add metadata:
- hook
- target audience
- CTA type
- related pillar
- asset status
- confidence level / verification needs

Wednesday — Approve and schedule

Human reviews tone, specificity, claim accuracy, and brand fit
Revise weak hooks or “AI-sounding” phrasing
Queue approved posts into scheduler / API layer

Thursday — Publish, engage, and harvest follow-ups

Reply to comments
Capture questions and objections
Convert conversations into future content angles

Friday — Review performance and learn

Review:
- comments
- saves
- profile visits
- inbound DMs
- booked calls
- topic resonance
Feed learnings back into prompts, templates, and pillar priorities

6. Recommended technical architecture

Option A — Recommended first pilot

Claude Code + Postiz + source-content repository

Best for: AI-native, API-oriented, self-hostable, lean operations

Suggested flow

Store source ideas and brand context in a structured repository (Notion, markdown repo, or simple DB)
Claude Code reads source materials and generates drafts
Human reviews drafts
Approved drafts are pushed to Postiz via API/CLI
Postiz schedules and publishes
Performance data and engagement notes are written back into the content system

Pros

Strong AI-native fit
API/CLI support
Self-hosting option
Broad platform support
Good match for future orchestration

Cons / caveats

Requires testing for reliability
Postiz API rate limits exist
Analytics/reporting fit may need supplementation
LinkedIn edge cases need validation

Option B — Simpler operational stack

Claude Code + Buffer

Best for: simpler, more established operational experience

Pros

More established brand
Simpler mental model

Cons / caveats

API transition / beta status is a meaningful uncertainty
Less explicitly “agentic”
May be less attractive if self-hosting or deeper orchestration matters

Option C — Collaboration-heavy workflow

Claude Code + Planable

Best for: strong approvals, previews, and multi-stakeholder review

Pros

Strong approval workflow emphasis
Better fit if multiple reviewers are involved

Cons / caveats

Less “AI-native CLI” oriented
Better as workflow hub than as programmable automation core

Option D — Broader all-in-one scheduler/analytics

Claude Code + Metricool

Best for: simpler operational management if publishing/reporting is the priority

Pros

Scheduling + analytics coverage
Likely easier for nontechnical users

Cons / caveats

Workflow fit for an AI-native agency still needs hands-on validation
API and tier details need direct confirmation

Tools, Resources, Links, and References

Primary / official sources

Anthropic — Claude Code with Chrome docs
https://code.claude.com/docs/en/chrome
Anthropic Help Center — Get started with Claude in Chrome
https://support.claude.com/en/articles/12012173-get-started-with-claude-in-chrome
LinkedIn — Posts API
https://learn.microsoft.com/en-us/linkedin/marketing/community-management/shares/posts-api?view=li-lms-2026-02
LinkedIn Business — Pages best practices
https://business.linkedin.com/advertise/linkedin-pages/best-practices
X — Create or Edit Post
https://docs.x.com/x-api/posts/create-post
X — API introduction
https://docs.x.com/x-api/introduction
Meta — Instagram content publishing
https://developers.facebook.com/docs/instagram-platform/content-publishing/
Google — YouTube Data API upload guide
https://developers.google.com/youtube/v3/guides/uploading_a_video
Google — YouTube Data API video resource docs
https://developers.google.com/youtube/v3/docs/videos
Postiz CLI docs
https://docs.postiz.com/cli/introduction
Postiz site / agent page
https://postiz.com/agent
https://postiz.com/
Postiz GitHub repository
https://github.com/gitroomhq/postiz-app
Postiz Public API docs
https://docs.postiz.com/public-api/introduction
Buffer developer docs
https://buffer.com/developers/api
Buffer developer API beta page
https://buffer.com/developer-api
Planable approval workflow positioning
https://planable.io/blog/social-media-approval-process/
Metricool API & integrations help
https://help.metricool.com/en/category/api-integrations-dc0snw/
Metricool main site
https://metricool.com/

Practitioner / user-reported sources (use carefully)

Reddit — B2B marketing thread on LinkedIn/personal profile use
https://www.reddit.com/r/b2bmarketing/comments/1qg4uzd/launching_a_b2b_product_in_2026_is_linkedin_still/
Reddit — Content marketing thread about company pages vs personal visibility
https://www.reddit.com/r/ContentMarketing/comments/1qp6yv7/company_pages_on_linkedin_are_basically_a/
Reddit — Another branding/LinkedIn discussion about personal brand vs company page
https://www.reddit.com/r/branding/comments/1qbpxsr/is_linkedin_actually_worth_it_for_b2b_lead_gen_or/
Reddit — Social media tools discussion favoring a smaller core stack
https://www.reddit.com/r/SocialMediaMarketing/comments/1r3ye0o/what_social_media_tools_are_actually_worth_using/
Reddit — Claude as social media drafting assistant, not full auto-poster
https://www.reddit.com/r/ClaudeAI/comments/1rmhayt/claudeai_as_social_media_manager/
Reddit — OpenClaw + Postiz automation workflow
https://www.reddit.com/r/AiAutomations/comments/1rn6geh/i_automated_my_entire_social_media_posting_with/
Reddit — Automation thread mentioning Claude Code/OpenClaw + approval loop
https://www.reddit.com/r/automation/comments/1rn6iwv/how_i_automated_posting_across_4_social_media/

User-provided style reference

Uploaded file: My Style Voice for AI.md
Internal project reference summarizing Dmitri’s preferred tone, paragraph length, anti-hype preferences, and later-stage grounded writing voice.

Risks, Caveats, and Red Flags

1) Fully autonomous posting is not the safest default

Verified / inferred from official capability limits: Claude Code browser use is powerful but still beta and risk-bearing.[1][2]
Tentative / speculative: UI automation can break on login expiration, CAPTCHAs, DOM changes, popups, and unexpected editor behavior.
Action: Keep a human approval gate and a scheduler/API layer.

2) LinkedIn integration complexity is real

Verified: LinkedIn permissions differ for organization vs member posting and some read permissions are restricted.[3]
Action: Validate the exact LinkedIn posting path Solanasis needs before building tooling around it.

3) YouTube API audits can surprise teams

Verified: Unverified YouTube API projects uploading videos default to private until audit.[9]
Action: Plan for this before assuming full autopublish to public YouTube videos.

4) Postiz is promising, but “best” is not verified

Verified: Postiz has strong docs and official positioning.[10][11][12][13]
Tentative / speculative: Reliability, analytics quality, and edge-case handling still need real-world testing for Solanasis.
Action: Run a small pilot before standardizing.

5) Buffer’s API state is in transition

Verified: Buffer says it is rebuilding its public API and that the new developer API is in beta.[14][15]
Action: Do not assume documentation or endpoints are stable without direct account-level confirmation.

6) Reddit evidence is directional, not definitive

Verified: Reddit threads were found and cited.
Tentative / speculative: These are useful practitioner signals, not authoritative truth.
Action: Treat Reddit as “field notes,” not proof.

7) Tool sprawl can quietly kill consistency

Tentative / speculative: The most likely failure mode is not lack of AI power; it is too many tools, unclear approval paths, and poor feedback loops.
Action: Start with one scheduler, one source-content system, one review loop.

8) Voice drift is a real problem

User-stated: Dmitri cares about authentic voice and dislikes bloated, sensational, generic AI writing.
Tentative / speculative: Without careful prompting and review, AI-generated content will quickly flatten into generic B2B content sludge.
Action: Build a voice profile, examples library, and anti-pattern checklist into the workflow.

Open Questions / What Still Needs Verification

Which scheduler should Solanasis actually adopt first?
Postiz is the best conceptual fit, but the decision should follow a hands-on pilot comparing:
- LinkedIn reliability
- Instagram/Reels support
- analytics usefulness
- approval UX
- API ergonomics
- failure handling
What exact LinkedIn integration path is required for Solanasis?
Do we need:
- personal-profile publishing only,
- company-page publishing only,
- both,
- or just scheduler-based posting?
How much analytics feedback can be pushed back into the content system automatically?
The earlier discussion mentioned feedback loops, but specific API/reporting implementation was not verified.
Does Metricool’s useful API access require a specific paid tier?
This was suggested but not fully verified from a primary source in this pass.
What are the best current workflows for pulling platform performance data back into Claude Code prompts?
The discussion implied this is valuable, but the best concrete implementation pattern remains open.
How much of Solanasis demand generation should be organic social vs founder outreach vs partnerships vs email?
This discussion focused on content, but an integrated GTM system still needs cross-channel design.
Which audience segment should the content strategy prioritize first?
Possible segments include:
- SMB founders/operators
- nonprofit executive directors / ops leaders
- RIAs / wealth-management-adjacent firms
- special-assets/foundation ecosystem
- IT / operations leads inside growing organizations
How much video capacity does Solanasis actually have?
YouTube is strategically attractive, but operational feasibility needs confirmation.

Suggested Next Steps

Immediate next steps (recommended)

Build a 90-day pilot around LinkedIn + X + light Instagram support.
Choose one scheduler to test first:
- Start with Postiz if AI-native / self-hosted / API-first matters most.
- Otherwise test Planable or Buffer depending on workflow priorities.
Create a voice and anti-voice brief for Solanasis content using Dmitri’s style reference.
Build a source-content repository:
- raw ideas
- call notes
- field observations
- reusable hooks
- approved terminology
- banned phrasing
Define 5–10 flagship post templates for LinkedIn and X.
Design a weekly approval rhythm with as little friction as possible.
Start tracking:
- post topic
- platform
- content pillar
- format
- comments
- saves
- DMs
- booked conversations
Decide whether YouTube is included in Phase 1 or Phase 2.
Run a direct technical validation of LinkedIn + scheduler + approval flow before committing.

Optional next artifact to create

A follow-on artifact should probably be:

“Solanasis 90-Day Content Operating System + Agent Prompt Pack”

That artifact should include:

exact weekly cadence
post templates
prompt library
approval SOP
metrics sheet
workflow diagrams
scheduler/API implementation plan
platform-specific content examples in Dmitri’s voice

Handoff Notes for Another AI

You are continuing work on Solanasis’s content strategy and AI-native social operations. Preserve the following context:

What Solanasis is

Founder-led B2B services firm
Themes: security, resilience, operations
Services likely include operational cleanup, resilience assessments, security work, systems integration, migrations, and responsible AI enablement
Brand should feel credible, practical, and sharp — not generic, fluffy, or hypey

What the user wants

A real 2026 content strategy
Heavily LinkedIn-centered
Also X and Instagram
Possibly one more high-value channel
AI-native workflows
API-compatible systems wherever possible
Real-world playbooks, not theoretical marketing fluff

Strongest current recommendations

LinkedIn first, led by Dmitri’s personal profile
Company page active but supporting
X second
Instagram support only
YouTube as the best additional compounding channel
Claude Code used upstream of publishing
Postiz as the first scheduler/API layer to test
Keep human review and engagement in the loop

What is verified vs not

Verified: Claude Code Chrome capabilities, LinkedIn/X/Instagram/YouTube publish support, Postiz CLI/API existence, Planable approval emphasis, Buffer beta/rebuild status
Not fully verified: best scheduler choice for Solanasis, Metricool tier specifics, exact analytics feedback architecture, whether Postiz is operationally best after real testing

Important style constraints

Dmitri dislikes overlong, generic, sensational AI copy
Keep paragraphs tight
Stay grounded, direct, warm, and practical
Prefer specificity over abstraction
Strong founder/operator tone is good; fake-polished agency tone is bad

Good next move for another AI

Use this document to produce:

a 90-day content OS,
platform-specific post templates,
a tool selection scorecard,
and a concrete implementation architecture for the chosen publish stack.

Reviewer Notes and Improvements Made

Reviewer availability

Verified: No dedicated reviewer-agent capability was available in this workflow.
Verified: A serious self-review pass was completed instead.

Improvements made over the original discussion

Separated facts from recommendations.
The original discussion blended technical facts, practitioner heuristics, and recommendations. This artifact separates them.
Verified core technical claims.
The following were checked against primary or high-confidence sources:
- Claude Code with Chrome capabilities
- LinkedIn posting permissions
- X API post creation
- Instagram content publishing support
- YouTube upload support and audit caveat
- Postiz CLI/API existence and compliance claims
- Buffer API beta/rebuild status
- Planable approval workflow positioning
Flagged weak or user-reported evidence.
Reddit-based claims were preserved but clearly labeled as practitioner reports, not authoritative fact.
Added missing operational concerns.
Missing considerations were added around:
- approval friction
- analytics feedback loops
- voice drift
- token security
- YouTube API audit constraints
- tool sprawl
- technical validation before lock-in
Turned the discussion into a usable operating model.
The original answer was strategy-forward; this version adds a practical weekly workflow, role model, channel roles, and implementation paths.

Optional Appendix — Structured YAML-Style Summary

document:
  title: "Solanasis Content Creation, Publishing, and AI-Native Social Ops Playbook"
  date: "2026-03-18"
  purpose:
    - guide
    - playbook
    - briefing_memo
    - AI_handoff_document
 
core_recommendation:
  status: tentative
  summary: >
    Use Claude Code upstream for research, drafting, and repurposing;
    use an official API or scheduler for publishing; keep a human in the loop
    for approvals and engagement.
 
channel_priority:
  - channel: LinkedIn
    role: primary
    status: tentative
    note: "Founder-led profile first, company page second/supporting"
  - channel: X
    role: secondary
    status: tentative
  - channel: Instagram
    role: support
    status: tentative
  - channel: YouTube
    role: compounding_source_channel
    status: tentative
 
tools:
  claude_code_chrome:
    status: verified
    best_use:
      - research
      - drafting
      - browser-assisted workflows
      - data extraction
    caution:
      - beta
      - visible browser window
      - inherent risks
  postiz:
    status: verified_exists
    recommended: true
    reasons:
      - api_first
      - cli_for_agents
      - self_hostable
      - broad_platform_support
    caveats:
      - needs_real_world_validation
      - api_rate_limits
  buffer:
    status: verified_exists
    caveats:
      - api_rebuild
      - beta
  planable:
    status: verified_exists
    strengths:
      - approvals
      - previews
      - collaboration
  metricool:
    status: partially_verified
    strengths:
      - scheduling
      - analytics
    open_question:
      - exact_api_tier_requirements
 
evidence_classes:
  verified:
    - claude_code_browser_capabilities
    - linkedin_posts_api_permissions
    - x_post_creation_api
    - instagram_content_publishing_support
    - youtube_upload_api_support
    - youtube_unverified_project_private_default
    - postiz_cli_api_support
    - postiz_official_oauth_no_scraping_claim
    - buffer_api_beta_rebuild_status
    - planable_approval_workflow_positioning
  user_stated:
    - linkedin_primary_goal
    - ai_native_agency_goal
    - api_access_preference
    - Dmitri_voice_preferences
  tentative:
    - founder_led_linkedin_first_model
    - youtube_as_best_extra_channel
    - postiz_as_best_first_tool_to_test
    - instagram_as_support_not_primary_leadgen
    - reddit_as_listening_channel

Solanasis Docs

Explorer

solanasis-content-creation-posting-strategy-research-playbook-handoff-2026-03-18

Solanasis Content Creation, Publishing, and AI-Native Social Ops Playbook — Research-Grade Extraction, Verification, and Handoff (2026-03-18)

Executive Summary

Purpose of This Document

Evidence Status Legend

Discussion Context

User goals and constraints

Brand / voice / operating context preserved for future use

Key Facts and Verified Findings

1) Claude Code can assist with browser-based content operations, but it is not the ideal final publishing backbone

2) Official platform APIs exist for core publishing use cases, but their constraints differ by platform

LinkedIn

X

Instagram

YouTube

3) The strongest “AI-native” publishing/scheduling fit discussed was Postiz, but it should still be treated as a tool choice, not a settled fact

4) Buffer, Planable, and Metricool remain relevant, but for different reasons

Buffer

Planable

Metricool

5) The most defensible channel hierarchy for Solanasis is LinkedIn first, X second, Instagram support, YouTube compounding

LinkedIn first

X second

Instagram as support

YouTube as the additional “compounding” channel

6) Agency and practitioner workflow evidence supports a smaller core stack, approvals, and human review — not maximal tool sprawl

Major Decisions and Conclusions

A. Recommended strategic direction

B. Recommended technical direction

C. Recommended operating principle

Reasoning, Tradeoffs, and Why It Matters

1) Why founder-led beats “faceless brand” for Solanasis

2) Why AI should sit upstream of publishing

3) Why one extra channel should be YouTube, not “everything else”

4) Why approval workflows matter more than “best AI model”

Recommended Playbook / Process

1. Operating model

Core principle

Roles

2. Channel strategy

LinkedIn

X

Instagram

YouTube

Reddit

3. Content pillars for Solanasis

Pillar 1 — Operational resilience reality

Pillar 2 — Security without theater

Pillar 3 — Messy systems into clean operations

Pillar 4 — Responsible AI for real operators

Pillar 5 — Founder/operator perspective

4. Content formats

LinkedIn

X

Instagram

YouTube

5. Weekly workflow

Monday — Source capture and prioritization

Tuesday — Draft and repurpose

Wednesday — Approve and schedule

Thursday — Publish, engage, and harvest follow-ups

Friday — Review performance and learn

6. Recommended technical architecture

Option A — Recommended first pilot

Option B — Simpler operational stack

Option C — Collaboration-heavy workflow

Option D — Broader all-in-one scheduler/analytics

Tools, Resources, Links, and References

Primary / official sources

Practitioner / user-reported sources (use carefully)

User-provided style reference

Risks, Caveats, and Red Flags

1) Fully autonomous posting is not the safest default

2) LinkedIn integration complexity is real

3) YouTube API audits can surprise teams

4) Postiz is promising, but “best” is not verified

5) Buffer’s API state is in transition

6) Reddit evidence is directional, not definitive