# Post Radar > Acts on engagement with monitored LinkedIn profiles. Qualifies commenters or reactors, checks fit, and drafts contextual follow-up tied to the actual post. - **Type:** Skill - **Install:** `agentstack add skill-swan-gtm-gtm-skills-post-radar` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [swan-gtm](https://agentstack.voostack.com/s/swan-gtm) - **Installs:** 0 - **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [swan-gtm](https://github.com/swan-gtm) - **Source:** https://github.com/swan-gtm/gtm-skills/tree/main/skills/ido-goldberg/post-radar ## Install ```sh agentstack add skill-swan-gtm-gtm-skills-post-radar ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About ## Instructions **Setup state.** Not yet configured for this org. Load the `Setup` sub-page and walk the user through wiring the `LINKEDIN_ENGAGEMENT` trigger with the right set of monitored profiles (founder, execs, company page, competitor pages, thought leaders), engagement types, noise filters, and follow-up sequence before running this play. (After setup is performed, rewrite this paragraph via `swan-update-skill` to describe the current state — monitored profile list, engagement types tracked (comments / reactions / reposts), noise filters applied, sequence wired, success metric, and last-refreshed date — so future runs see the current configuration without re-checking.) ### When this fires A `LINKEDIN_ENGAGEMENT` trigger lands. Payload includes: the post URL, the post content (or summary), the engager's name + LinkedIn URL, engagement type (comment / reaction / repost), and the comment text if it's a comment. Speed matters. Replying within 24 hours converts much better than 72+ hours. ### Step 1 — Filter the noise Not every engagement is worth acting on. Quick filters: - **Generic emoji-only reactions** → low signal, skip unless they're a high-value account - **One-word comments** ("Love this!", "100%") → skip unless from an ICP-fit person - **Competitors and vendors** → never engage as a sales motion; flag for awareness only - **Existing customers** → route to CSM, not to outbound - **Internal employees** → skip Set a minimum threshold: comments ≥ ~15 words, or any engagement from a known ICP company. ### Step 2 — Enrich the engager `swan-enrich-contact` with the LinkedIn URL. Returns: title, company, tenure, role family. Cheap and necessary. The engager's title + company are the two things that determine the right reply. ### Step 3 — ICP and CRM check in parallel Two checks: 1. **ICP fit** — does the engager's company match the org's ICP? Use `swan-search-companies` for the domain; if not in Swan, briefly check firmographic match. 2. **CRM relationship** — `hubspot-search-objects` (contacts, filter by email or LinkedIn URL). Is this person already in the system? Is there an active deal or owner? Outcome routing: - ICP fit + no relationship → cold outreach play, draft message - ICP fit + active relationship → route to owner with a "they engaged on your post" alert - Non-ICP but interesting → flag for awareness, no outreach - Non-ICP and noise → discard ### Step 4 — Read the post they engaged with The engager interacted with specific content. The reply must reference what they actually engaged with — generic "saw your comment" outreach is worse than ignoring them. `swan-fetch-scraped-url` on the post URL if the trigger payload didn't include the full post content. Otherwise use what's in the payload. Extract: the post's topic, the angle of their comment / reaction, any specific question or statement they made. ### Step 5 — Choose the response channel LinkedIn engagement begets LinkedIn response. Options: - **LinkedIn comment reply** — public, low-commitment, opens a conversation - **LinkedIn DM / connection request** — direct, slightly more aggressive - **Email follow-up** — if the relationship warrants and the email is known, fine to escalate Default is LinkedIn comment reply first, then DM if they respond. ### Step 6 — Draft the response Templates: **Reply to a substantive comment:** > "[Engager], totally agree on [their specific point]. We see the same with [your customers] — usually it shows up as [specific manifestation]. Curious how you're thinking about [their stated challenge]." **DM after a reaction-only signal (lower priority):** > "Hey [name] — noticed you reacted to [post topic]. We work with [persona] on [adjacent problem] — would love to hear how you're approaching it at [their company]." Keep it short. Sales-y openings kill the warmth of the signal. ### Step 7 — Route or send For unowned ICP-fit prospects: hand off to `reach-out` or call `swan-build-sequence` (single-touch LinkedIn-first, no aggressive cadence). For owned accounts: `hubspot-create-task` for the owner with the post URL, the engager, and a suggested message. Don't auto-send into a rep's territory. ### Step 8 — Log `swan-update-company` to tag the engager's company with the engagement signal. `swan-save-message-example` if the message style was approved — feed the engagement-reply voice back into the org's tone library. ### Rules - MUST reply within 24 hours to retain the warmth advantage. - MUST reference the specific post and the engager's specific action. Generic outreach negates the signal. - MUST check ICP fit and CRM ownership before drafting. - NEVER pitch in the LinkedIn comment thread. Use the comment to start a conversation; pitch in DM or email. - NEVER message every reaction. Filter aggressively to substantive comments + ICP fit. - NEVER reply to competitor or vendor engagement as a sales motion. - If the engager has engaged on multiple posts in the last 90 days, treat as a hot signal — escalate priority. `swan-linkedin-social-media-presence` on the engager can surface this. - If a tool result is truncated, read from `files/tool-outputs/_.json` in `swan-execute-code`. ### Tighten over time After 10-20 fires, read recent reply-rate data via `swan-search-sequences` filtered to this play's executions. Identify which monitored profiles produced highest-converting engagers (often founder + one named exec dominate). Recommend the user drop low-yield monitored profiles, raise the minimum comment-length threshold, or restrict to ICP-fit companies only before the next batch. ## Source & license This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [swan-gtm](https://github.com/swan-gtm) - **Source:** [swan-gtm/gtm-skills](https://github.com/swan-gtm/gtm-skills) - **License:** MIT Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/skill-swan-gtm-gtm-skills-post-radar - Seller: https://agentstack.voostack.com/s/swan-gtm - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.