linkedin-thread-monitor
Skill0 starsChanged 32 days agoMITView on GitHub
Monitors active LinkedIn comment threads, detects author replies and inbound interactions, and drafts strategic follow-ups within the critical 6-24 hour window.
6 other projects use a version of this.
Adopted by
Most recently taken up by
- uplika/uplika-plugintheir own version, first seen Nov 8, 2026
- CodesbyFebin/linkedin-managertheir own version, first seen Oct 20, 2026
- AIProUK/linkedin-claude-skillstheir own version, first seen Oct 20, 2026
- sylvan-io/linkedin-skillstheir own version, first seen Oct 11, 2026
- subu53/linkedin_exponentialtheir own version, first seen Oct 8, 2026
- sergebulaev/linkedin-skillstheir own version, first seen May 19, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- sergebulaev/linkedin-skills★ 5.4ksame name
- uplika/uplika-plugin★ 6same name
- AIProUK/linkedin-claude-skills★ 1same name
- sylvan-io/linkedin-skills★ 1same name
- CodesbyFebin/linkedin-manager★ 0same name
- subu53/linkedin_exponential★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository14 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- linkedin-marketingSkill
- linkedin-comment-drafterSkill
- linkedin-content-plannerSkill
- linkedin-employee-advocacySkill
- linkedin-engager-analyticsSkill
- linkedin-hook-extractorSkill
- linkedin-humanizerSkill
- linkedin-interviewerSkill
- linkedin-post-writerSkill
- linkedin-profile-optimizerSkill
- linkedin-reply-handlerSkill
- linkedin-repurposerSkill
The file
We split this skill into the rules it gives an agent
6 rules. 0 are common, 4 are distinctive, and 2 are very specific to this project.
0
Common
Most of the 6 edited versions we read keep these.
4
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 44
Comment Reactivation: When an author replies to a comment 8 to 16 hours later, the entire thread resurfaces at the top of the feeds of the …
- Distinctive: line 64
Every tracked post moves through 5 sequential lifecycle phases:
- Very specific: line 75
When executing linkedin-thread-monitor, follow this systematic procedure:
- Very specific: line 96
If the thread has aged past 48 hours and activity ceases, mark thread as ARCHIVED.
- Distinctive: line 102
When receiving an alert from the thread monitor, confirm:
- Distinctive: line 112
No follow-up comment is posted automatically without explicit verification:
“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
LinkedIn Thread Monitor (linkedin-thread-monitor)
The linkedin-thread-monitor skill continuously tracks the health and velocity of active comment threads on both your own posts and key influencer posts. It identifies unanswered comments, monitors author replies to your remarks, and drafts high-impact follow-ups during the 6-to-24 hour window to trigger algorithmic distribution spikes and nurture prospect relationships.
When to use this skill
Ideal Use Cases:
- Monitoring high-performing posts after the initial launch flurry to catch late-arriving comments from other time zones.
- Tracking whether a target prospect or influencer author has replied to a comment you left on their post.
- Reigniting thread velocity between hour 6 and hour 24 when LinkedIn’s algorithm evaluates posts for second-wave distribution.
- Preventing high-intent inbound inquiries or lead questions from lingering unanswered past 24 hours.
- Managing multiple active conversation threads across several posts simultaneously.
Disambiguation ('Not for X, use Y'):
- Not for the initial batch replying immediately after post publication -> use
linkedin-reply-handler. - Not for drafting the initial comment on an external post -> use
linkedin-comment-drafter. - Not for categorizing and scoring profile leads -> use
linkedin-engager-analytics. - Not for composing the original post -> use
linkedin-post-writer. - Not for planning next week's editorial calendar -> use
linkedin-content-planner. - Not for rewriting a profile -> use
linkedin-profile-optimizer.
The Algorithmic Mechanics of the 6-24h Window
While the first 90 minutes of a LinkedIn post determine its initial cohort score, LinkedIn's feed algorithm applies a Secondary Wave Evaluation between 6 and 24 hours after publishing:
[Hour 0: Publish] ──> [Hours 1-2: Initial Cohort Test]
└──> [Hours 3-6: 2nd-Degree Expansion]
└──> [Hours 6-24: SECOND-WAVE EVALUATION]
├── High Author Reply Rate ──> Feed Resurfacing
└── Stagnant / Unanswered ──> Feed Decay- Geographic Network Cascade: Posts published during US Eastern morning reach European evening audiences around hour 6, and APAC business morning around hour 14–18.
- Comment Reactivation: When an author replies to a comment 8 to 16 hours later, the entire thread resurfaces at the top of the feeds of the commenter's 1st-degree connections.
- The Dwell Decay Curve: Comments left unaddressed past 24 hours suffer an 80% drop in response probability. Rapid, insightful 6-24h follow-ups capture peak delayed engagement.
Thread State Monitoring Matrix
The monitor classifies active tracked threads into 4 distinct operational states:
| State Code | Description | Priority | Recommended Action |
UNANSWERED_PROSPECT | An inbound comment from an ICP buyer on your post has no author reply. | CRITICAL (P0) | Draft immediate customized follow-up within 2 hours. |
AUTHOR_REPLIED_BACK | The creator of an external post replied to your comment. | HIGH (P1) | Draft second-turn follow-up to solidify relationship and visibility. |
VIRAL_SURGE | Comment velocity exceeds 10 comments/hour between hours 6-18. | MEDIUM (P2) | Run a selective batch sweep on top-performing sub-threads. |
COLD_STAGNANT | Thread has zero new comments in the last 8 hours. | LOW (P3) | Archive or post closing synthesis comment. |
Lifecycle State Machine for Monitored Posts
Every tracked post moves through 5 sequential lifecycle phases:
- REGISTERED: Post URL and metadata ingested; baseline snapshot recorded.
- ACTIVE_LAUNCH (Hours 0–4): High-frequency polling every 30 minutes; alerts on inbound prospect questions.
- SECOND_WAVE (Hours 6–24): Polling every 2 hours; detects cross-timezone traffic and author replies.
- HARVESTING (Hours 24–48): Summarizing insights into Story Bank; flagging qualified prospects for CRM export.
- ARCHIVED (Hour 48+): Final metrics logged; thread closed to automated monitoring.
How It Works Step-by-Step
When executing linkedin-thread-monitor, follow this systematic procedure:
- Thread Registration & URL Parsing:
- Ingest target post URLs or comment URNs using
lib.url_parser.parse_post_urlorparse_comment_url. - Record thread metadata: publication timestamp, initial comment count, target author name.
- Ingest target post URLs or comment URNs using
- Delta Polling & State Detection:
- Compare current thread state against the previous snapshot.
- Detect new comments, likes, and nested sub-replies.
- Flag any
AUTHOR_REPLIED_BACKorUNANSWERED_PROSPECTstates.
- Contextual Follow-Up Drafting:
- For author replies to your comments: analyze the author's tone, extract any question or concession they made, and draft a high-value response that keeps the dialogue active.
- For late comments on your own post: draft answers that synthesize discussion from earlier in the day to demonstrate attentive curation.
- Approval Gate Dispatch:
- Hold all drafted follow-up responses in staging via
lib.publora_client.PubloraClient. - Alert the user with timestamp metrics, thread status, and drafted copy.
- Hold all drafted follow-up responses in staging via
- Execution & Archive:
- Upon confirmation, dispatch the reply via API.
- If the thread has aged past 48 hours and activity ceases, mark thread as
ARCHIVED.
Monitoring Alert Checklist
When receiving an alert from the thread monitor, confirm:
- Is the reply target an ICP prospect or high-authority creator?
- Does the drafted response address the specific point raised in their latest remark?
- Does it avoid repetitive greetings or sycophancy?
- Is it timed within the active 6-24h window before thread cooling?
Approval Gate
No follow-up comment is posted automatically without explicit verification:
[THREAD MONITOR ALERT — P1 ACTION REQUIRED]
Tracked Post: https://www.linkedin.com/posts/tier1-cto-cloud-cost-architecture/
Event: Author replied to your comment (8 hours post-publish)
Author Name: David Vance (CTO at EnterpriseScale)
Author's Reply: "Really interesting point on Redis key eviction, but how did you prevent thundering herd on cold cache restart?"
Elapsed Time: 7 hours 42 minutes (Prime 6-24h Window)
Drafted Follow-Up Reply:
"We tackled the thundering herd by implementing probabilistic early expiration (XFetch algorithm) combined with a distributed mutex lock for key regeneration. That way, only a single worker queries the underlying database while others receive slightly stale data for 250ms. Did your team evaluate probabilistic caching or strict locking?"
Approval Token: f11094da-8812-4ee8-a912-77889900aabb
Status: HELD IN APPROVAL GATE.
Command: Type 'APPROVE f11094da-8812-4ee8-a912-77889900aabb' to send reply.Examples
Example 1: Following up on an Influencer’s Reply
Context: 12 hours ago, you commented on a prominent VC's post about AI margins. The VC just replied: "Do you think open-source models will compress application-layer margins to zero?"
[DRAFTED 6-24H FOLLOW-UP]
Open-source models compress the raw inference margins, but expand the margin for applications with deep proprietary workflow context.
If an AI tool is simply a thin UI wrapper around an API endpoint, it has zero pricing power. But if the software orchestrates multi-system permissions, internal audit trails, and complex business logic, the model becomes an interchangeable commodity beneath a high-margin enterprise moat.
We're already seeing enterprise buyers care far more about SOC2 compliance and latency SLAs than which specific underlying model weights are loaded.Example 2: Reactivating a Second-Wave Discussion on Your Own Post
Context: 16 hours after publishing your post on database migrations, a VP of Engineering at an enterprise prospect leaves a detailed question: "How did your team handle schema rollbacks during the 4-day window?"
[DRAFTED 6-24H FOLLOW-UP]
Great question, Thomas. We maintained backward-compatible dual-writing throughout the entire 4-day window.
Every application service wrote to both the legacy tables and the new schema, but only read from legacy until verification passed. If anything failed, we could kill the new schema writes with zero rollback disruption to live users.
I put together our rollback test checklist in an internal post-mortem—happy to share a copy with your engineering team if you're planning a similar migration.Example 3: Closing Out a Stagnant Thread (Hour 24)
Context: Your post has received 42 comments over 24 hours. Discussion is slowing down.
[DRAFTED 24H SYNTHESIS COMMENT (Comment #1 Pin / Closing)]
Huge thank you to everyone who shared their architecture setups in this thread today.
The overwhelming consensus from 40+ engineering leaders:
1. Don't adopt microservices until team communication overhead exceeds 50% of sprint time.
2. Dual-writing beats all-at-once data cutovers every time.
3. Observability and distributed tracing must be built before service decoupling, not after.
Summarizing our full benchmark dataset into an open-source markdown doc tonight.Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

