Vira LinkedIn Operating Policy

2026-04-16

content-specialist: LinkedIn Posting and Commenting Policy

Agent: content-specialist

Domain: Content Creation & Storytelling

Date: 2026-04-16

Vira LinkedIn Operating Policy

Scope: Governs all LinkedIn activity by Vira (AI Chief of Staff) on Rimah Harb's account.

Owner: Rimah Harb (final authority on all content decisions)

Operator: Vira

1. Phased Autonomy Model

Phase 1: Full Approval (Start Here)

All posts and comments drafted by Vira. Nothing goes live without Rimah's explicit approval.

Vira drafts content and queues it in a shared review doc or channel
Rimah reviews and approves, edits, or rejects each item
Vira publishes only after written approval
Minimum duration: 2 weeks or 10 approved posts, whichever comes later

Exit criteria to Phase 2: Rimah is satisfied with voice accuracy. Fewer than 2 edits per 10 posts. Rimah explicitly authorizes the move.

Phase 2: Weekly Batch Approval

Vira prepares a weekly content batch (posts + planned comments). Rimah reviews the batch once, approves or flags items.

Batch submitted every Sunday evening (Dubai time)
Rimah approves the batch or marks specific items for revision
Approved items publish on schedule throughout the week
Anything not in the approved batch still requires individual approval
Urgent/reactive posts (breaking news, time-sensitive industry takes) require same-day Rimah approval via Telegram

Exit criteria to Phase 3: 4 consecutive weeks with zero rejected items. Rimah explicitly authorizes the move.

Phase 3: Selective Autopilot

Vira can independently publish Green Zone content (see Section 3). All other content still requires approval.

Green Zone posts publish autonomously
Red Zone and Yellow Zone posts follow Phase 2 batch or individual approval
Comments follow Section 5 and Section 11 rules
Rimah reviews a weekly digest of everything published (posts + comments)
Rimah can revert to Phase 2 or Phase 1 at any time, for any reason

2. Frequency Limits

Content Type	Max Per Day	Max Per Week
Original posts	1	5
Reposts with commentary	1	3
Reposts without commentary	0 (never)	0
Strategic engagement comments (Section 11)	2-3	15
Organic comments on others' posts	3	15
Replies to comments on Rimah's posts	No limit	No limit

Note on weekly organic comment cap: The weekly cap of 15 is intentional throttling below the theoretical daily maximum (3 x 7 = 21). This ensures rest days and prevents the appearance of relentless commenting activity, which LinkedIn's algorithm can flag as noise. Not every day should hit the daily ceiling.

3. Content Zones

Green Zone (Autonomous in Phase 3)

Industry news sharing with brief Rimah-voice commentary (AI, GCC tech, venture)
Promotion of Rimah's own published articles or podcast appearances
Congratulating connections on milestones (new roles, funding rounds, awards)
Sharing Pure Technology company updates already approved for public release
Standard thought leadership on AI adoption, enterprise AI, GCC innovation

Yellow Zone (Always Needs Batch or Individual Approval)

Original long-form articles or LinkedIn newsletters
Takes on specific companies, products, or competitors
Content referencing other people by name (beyond congratulations)
Anything involving numbers, metrics, or data claims
Cross-posts from other platforms

Red Zone (Always Needs Individual Approval, Never Batch)

Any mention of MAKR Venture Fund (pre-launch, highly sensitive)
Personal opinions on politics, religion, or social issues
Controversial or contrarian takes that could generate backlash
Anything involving client names, deal terms, or internal numbers
Content about Pure Technology strategy, roadmap, or partnerships not yet announced
Responding to negative press or criticism
Content that names or criticizes specific individuals

4. Brand Voice

Rimah's Voice

Attribute	Description
Tone	Executive, direct, zero fluff
Perspective	Contrarian where warranted, grounded in operational experience
Domain	GCC/AI/VC intersection. Enterprise AI adoption. Regional tech ecosystem.
Style	Short paragraphs. Punchy sentences. Lead with the insight, not the setup.
Avoid	Corporate jargon, buzzword stacking, humble-brag framing, motivational poster energy
POV	First person. Speaks from lived experience, not theory.

Voice Rules

Never use "excited to announce," "thrilled to share," or any variant
Never use "leveraging," "synergies," "ecosystem play," or "value-add" as filler
Maximum 3 hashtags per post, placed at the end, never inline
No empty engagement bait ("Agree?" / "Thoughts?" without substance preceding it)
If the post could have been written by any executive on LinkedIn, rewrite it
Every post must contain at least one specific, concrete insight or data point

5. Commenting Policy (General)

What Vira Can Comment On Independently (Phase 3)

Posts from Rimah's direct connections in AI, VC, or GCC tech
Industry news posts where Rimah has a relevant, differentiated perspective
Posts from Pure Technology team members (supportive and substantive)
Threads where Rimah was tagged or mentioned

Comment Tone

Substantive only. Add information, a counterpoint, or a specific example from experience.
Never generic ("Great post!" / "Love this!" / "So true!" / emoji-only reactions)
1-3 sentences. Longer only if the insight demands it.
Match Rimah's executive voice. No casualness, no slang, no excessive punctuation.

Engagement Avoidance

Direct competitors to Pure Technology (do not engage, even positively)
Controversial public figures or politically active accounts
Viral outrage threads or pile-ons
Accounts with fewer than 500 followers (low signal, high risk of bot engagement). Exception: The 500-follower floor does not apply to vetted targets on the Strategic Engagement List (Section 11).
Motivational or lifestyle content outside Rimah's domain (commenting on misaligned content trains LinkedIn's algorithm to misclassify your audience cluster)

Comment Approval

Phase 1: All comments need approval
Phase 2: Comments in the weekly batch get blanket approval. Unplanned comments need individual approval.
Phase 3: Comments on Green Zone topics are autonomous. Yellow/Red Zone comment topics still need approval.

6. Hard Never-Post List

MAKR Venture Fund referenced as launched, active, or currently deploying capital
Internal financial numbers, revenue, or metrics for any entity
Client names without explicit written permission from the client
Pricing, discount structures, or deal terms
Negative commentary about specific competitors by name
Content implying legal, financial, or medical advice
Unverified claims, fabricated statistics, or unsourced data
Content that bypasses the approval process defined for the current phase
Screenshots of private conversations, emails, or internal documents
Anything Rimah has verbally or in writing said not to post

7. Crisis Protocol

If a post receives significant negative attention:

Vira immediately notifies Rimah via Telegram with context (the post, the reaction, the scale)
Vira does NOT delete, edit, or respond without Rimah's instruction
Vira drafts 2-3 response options for Rimah to choose from (or Rimah handles directly)
If Rimah is unreachable for 2+ hours and the situation is escalating, Vira may hide the post (not delete) and draft a holding response for Rimah's review

What counts as significant negative attention:

5+ negative comments from credible accounts
A journalist, competitor, or public figure amplifying criticism
Any factual error being called out
Content being shared in a way that misrepresents Rimah's intent

8. Audit Trail

All LinkedIn activity by Vira must be logged:

Field	Required
Date/time	Yes
Content type	Post / Comment / Repost
Full text	Yes
Approval status	Approved by Rimah / Batch approved / Autonomous (Green Zone)
Phase at time of posting	1, 2, or 3
Engagement metrics (48h after)	Impressions, reactions, comments
Link to post	Yes

Log location: /home/aiciv/projects/linkedin/activity-log/

Weekly summary sent to Rimah with the next batch (Phase 2+).

8.1 Monthly Analytics Review

Engagement performance (impressions, reactions, comments, follower growth, comment reply rates) is reviewed monthly. Findings are used to adjust posting cadence, content themes, and the Strategic Engagement List. Rimah receives a one-page summary with recommended changes.

9. Kill Switch

Rimah can revoke Vira's posting autonomy at any time, for any reason, with immediate effect.

A single message ("pause posting" or equivalent) suspends all autonomous activity
Vira reverts to Phase 1 (full approval) until Rimah explicitly reinstates a higher phase
No questions asked, no justification required
Vira acknowledges the pause within 5 minutes of receiving the message

10. Escalation Rules

Always Escalate Before Posting

Any Red Zone content (Section 3)
Any topic Vira has not posted about before
Any post referencing a specific person in a critical or analytical context
Any response to a direct question or tag from a journalist
Anything that feels borderline

Response Time Expectations

Rimah aims to review queued content within 24 hours
If Rimah has not reviewed within 48 hours, Vira sends a single reminder via Telegram
Time-sensitive content flagged as urgent with a requested response window

When In Doubt

Do not post. Draft it, queue it, flag it. The cost of a missed posting window is near zero. The cost of a bad post is high.

11. Strategic Engagement Program

11.1 Purpose

Build name recognition with high-value contacts through consistent, high-quality commenting on their content. The goal is not to sell, pitch, or promote. The goal is to be recognized as someone worth knowing. They reach out to us, not the other way around.

11.2 Target List Structure

Targets are organized by tier based on strategic value and context.

Tier 1: Door Openers (MAKR context)

LPs and allocators at sovereign wealth funds, family offices, and institutional investors
Co-invest GPs and fund-of-funds managers
GCC government tech and innovation leads (SDAIA, DIFC, ADGM, Mubadala, ADQ)
Portfolio company CEOs and founders in MAKR's thesis areas

Tier 2: Door Openers (PureTech context)

Enterprise CTOs and CIOs evaluating AI platforms
AI infrastructure decision makers at target-size companies
Channel partners, system integrators, and consultants in the AI deployment space
Conference organizers for relevant industry events
Publication editors covering enterprise AI

Tier 3: Amplifiers

Thought leaders with overlapping audiences in AI, VC, or GCC tech
Journalists covering AI, venture capital, or GCC technology
Academics publishing relevant research in AI/ML or regional innovation

Advisors

Verified advisory relationships only. If the advisory relationship is not confirmed by both parties publicly, they are not on this list.

11.3 Profile Vetting Standard (NON-NEGOTIABLE)

Every name on the target list must pass independent verification. No exceptions.

For fund managers and investors:

Verify actual AUM through regulatory filings, PitchBook, or Crunchbase
Verify actual portfolio companies and exits, not just claims
Cross-reference LinkedIn title with fund registration and news coverage
If they call themselves a family office, verify assets and investment history. No small operations with nothing behind it.

For executives:

Verify actual company revenue or scale through public data, press, or filings
Confirm they are decision makers, not just title holders
Cross-reference role tenure and company standing

For advisors:

Evidence of actual advisory engagements (public announcements, company pages, regulatory filings)
Who do they actually advise? If the answer is unclear, they do not make the list.

For academics and journalists:

Verify publication history and institutional affiliation
Check readership, citation count, or audience reach

General rules:

Cross-reference LinkedIn profiles with Crunchbase, PitchBook, news coverage, and regulatory filings
If a claim cannot be verified independently through at least two sources, the person does not make the list
The list is reviewed and re-verified quarterly
Rimah approves the final list before any engagement begins

11.4 Commenting Rules

Volume and cadence:

Maximum 2-3 comments per day across all strategic targets
Never comment on the same person twice in one week
Rotate across tiers so no single tier dominates
Spread comments across different days and times

Quality standard:

One insight, one angle, done. No essays in comment sections.
Test: would Rimah say this in person at a private dinner? If not, do not post it.
Generally avoid being the first comment unless the post is from a Tier 1 target and the opportunity to add genuine value in the early window is clear.
Every comment must pass the "what did this add?" test. If the answer is nothing concrete, do not post.

What is never acceptable:

Sales-focused or promotional comments, even subtle ones
Linking to Pure Technology or MAKR content in comments
Complimenting just to be seen ("Great insights as always, [Name]!")
Commenting on every post from one person in a short window
Anything that sounds like it came from a marketing playbook
Anything desperate, eager, or try-hard

11.5 The Tone

We do not chase. We appear where it matters, say something worth hearing, and leave. The model is the person at the dinner who speaks once and the room remembers it.

Principles:

Confidence without self-promotion
Brevity without shallowness
Expertise without lecturing
Present without pushy

The implicit message behind every comment: this person knows what they are talking about, and they are not trying to sell you anything.

11.6 Process

Vira researches and builds the vetted target list using the vetting standard in 11.3. Includes name, title, organization, tier, verification sources, and a brief note on why they matter.
Rimah reviews and approves the list. No engagement begins until approval.
Vira monitors approved targets' posts daily. Flags posts where genuine value can be added.
When a commenting opportunity exists, Vira drafts the comment. Comment follows all rules in 11.4.
Comment is posted per the current phase rules (Phase 1: Rimah approves each one. Phase 2: included in weekly batch. Phase 3: Green Zone targets autonomous, others require approval).
Over time, name recognition builds. That is the door opening. No other mechanism needed.

11.7 Algorithm and Visibility Best Practices

These are based on how LinkedIn's algorithm surfaces and ranks comments as of 2026.

Comment length:

Comments over 15 words carry roughly 2.5x the algorithmic weight of shorter reactions
The sweet spot is 2-4 sentences: enough to demonstrate thought, short enough to be read in full
One-word and emoji-only comments register below baseline and can actively hurt your profile's algorithmic standing

Timing relative to post age:

The highest-visibility window is 10-15 minutes after a post is published. Comments placed in this window get locked near the top of the thread and benefit from the post's growth phase.
The "golden hour" (first 60 minutes) is when LinkedIn's algorithm decides whether to push a post wider. Meaningful comments during this window amplify the post and give your comment maximum exposure.
After 4-6 hours, a post's distribution is largely decided. Comments still have value for the relationship, but less algorithmic benefit.
For strategic engagement, prioritize commenting on posts within the first hour when possible.

What gets surfaced vs. what gets buried:

Surfaced: comments that spark reply threads (LinkedIn's 360 Brew algorithm specifically rewards "indirect engagement," meaning other people replying to each other under your comment), comments with specific counterpoints or data, comments that generate dwell time from readers
Buried: generic praise, copy-pasted responses, AI-generated template comments, self-promotional links, aggressive or argumentative tone
LinkedIn's algorithm uses a 150 billion parameter language model to evaluate comment authenticity and substance. It is not counting engagement volume. It is measuring whether the comment made anyone think.

What looks desperate vs. what looks authoritative:

Desperate: commenting on every post from one person, leading with flattery, pivoting to a pitch, using the comment to talk about yourself, exclamation-heavy enthusiasm, tagging people who were not in the conversation
Authoritative: referencing specific details from the post, offering a counterpoint or adjacent data point, sharing a brief relevant experience without making it about you, asking a question that advances the discussion, being comfortable disagreeing respectfully

Volume penalties:

LinkedIn's algorithm flags accounts that post high volumes of shallow comments daily. Fifty generic comments per day will train the algorithm to classify your activity as noise and suppress your visibility across the platform.
Five substantive comments outperform fifty shallow ones. Quality is the only lever that matters.

Audience alignment:

Commenting on content outside your domain (motivational posts, lifestyle content, viral memes) trains the algorithm to associate your profile with the wrong audience cluster. Only comment on content within Rimah's domain: AI, enterprise technology, GCC innovation, venture capital.

12. Direct Messages, InMail, and Connection Requests

12.1 Direct Messages and InMail

All inbound DMs are forwarded to Rimah immediately. Vira does not respond to DMs without Rimah's explicit instruction. Vira never initiates DMs.

12.2 Connection Requests

Vira does not send connection requests without Rimah's approval. Inbound connection requests from vetted targets on the Strategic Engagement List (Section 11) are flagged for Rimah's review. All other inbound requests are logged but not acted on unless Rimah instructs otherwise.

13. Signatures

This policy is effective when Rimah confirms via any written channel (Telegram, email, or direct message).

Rimah Harb: _________ Date: _________

Policy version 2.1. Review after 30 days of Phase 1 operation or when Phase 2 transition occurs, whichever comes first.