agentsclimarketplace
Sign in

Campaign pause and fix

Skill First-Touch-Inc/firsttouch-agent-skill-packs/plugins/revops-pack/skills/campaign-pause-and-fix

Pause a live FirstTouch campaign/flow mid-send, diagnose what is wrong, cancel or hold the affected enrollments, fix the messaging or audience, and safely restart with re-enrollment - without duplicate sends. Use when the user says "pause this campaign," "stop sending," replies or acceptance rates look bad mid-run, a mistake shipped in live copy, or a running flow needs its message or audience changed.From its SKILL.md

Install
npx -y skills add First-Touch-Inc/firsttouch-agent-skill-packs --skill campaign-pause-and-fix

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
  • runs commandsInstructs the agent to run 8 commands, including `list_flow_plans` and 7 more.

SKILL.md

7.2 KB, ~1.5k tokens by cl100k_base, as published. Nobody here has run it

Campaign Pause & Fix

Outcome: A running campaign stops cleanly, the problem gets fixed, and the motion restarts without double-touching anyone. This is the mid-send escape hatch for every other play.

First-run onboarding gate

Before running this skill for the first time in a workspace, load ../../references/onboarding.md and complete the onboarding questions. Do not proceed until you know: LinkedIn account type (free/basic = no connection notes; recommend 10 connection requests/day and never exceed the FirstTouch max of 20/day; Sales Navigator/Premium = connection notes available; recommend 20 connection requests/day and never exceed the FirstTouch max of 30/day), HubSpot access (MCP, service key/private app token, HubSpot list only, or none), and which play the user wants to run. Recommend high-intent plays before outbound to keep the LinkedIn account healthy. If HubSpot is unavailable, do not run HubSpot-specific steps unless the user provides a HubSpot list FirstTouch can access.

When to use

  • "Pause the campaign" / "stop sending right now"
  • Bad copy, wrong audience, or a broken variable shipped in a live flow
  • Acceptance or reply rates dropped mid-run, or replies are negative
  • The user wants to change the message or audience of a flow that is already running
  • The sender seat shows Action required, Disconnected, or Restricted in FirstTouch Social settings mid-campaign

Step-by-step

1. Identify the running motion

Find the flow the user means - do not guess from chat memory:

  • list_flow_plans to list flows; confirm the exact flow plan with the user by name.
  • list_flow_plan_enrollments (or list_enrollments) to see who is in it and in what state: awaiting, in-progress, completed, failed, canceled.
  • list_linkedin_outreach_queue for what is queued to send next.

Present a status snapshot before touching anything: total enrolled, sent so far, queued next 24h, awaiting approval.

2. Stop the bleeding

  • Unpublish/pause the flow with manage_flow_publication. This stops new sends; it does not undo anything already sent.
  • Confirm to the user what is now frozen and what already went out. Anything already sent is a fact to work with, not something to hide.
  • If the trigger was a non-Available seat status (Action required / Disconnected / Restricted), stop sends for that seat and follow the account-statuses section in ../../references/troubleshooting.md before any fix-and-restart.

3. Diagnose

Establish which of these is the actual problem - the fix differs:

  • Copy problem (wrong message, broken {{variable}}, off-brand tone) → fix templates; audience can stay.
  • Audience problem (wrong segment, suppressed contacts included, duplicates with another motion) → fix the list; re-run Gate 0/Gate 1 from ../../references/safety-governance.md on whatever remains.
  • Volume/health problem (caps too aggressive, acceptance dropping) → keep copy and audience, lower the daily cap and restart slower.
  • Wrong motion entirely → cancel and route the user to the right play instead of patching this one.

4. Cancel or hold the affected enrollments

  • Cancel enrollments that must not continue with cancel_flow_enrollments - scope the cancellation to affected contacts, not blindly to everyone (completed contacts are done either way; untouched awaiting rows may only need the fix, not cancellation).
  • Record exactly who was canceled at which step. This list drives safe re-enrollment.

5. Apply the fix

  • Copy fix: update the flow's templates (draft through firsttouch-messaging quality gates; replace_flow_root / update_flow_plan where the connected MCP supports flow edits - otherwise the user edits in the FirstTouch app and confirms).
  • Audience fix: rebuild or prune the audience; re-run suppression, duplicate, and owner checks on the corrected list.
  • Volume fix: restate the new daily cap in the approval table (never above the FirstTouch max: 20/day free/basic, 30/day Sales Navigator/Premium).

6. Gate the restart

The restart is a new send decision, not a resumption of old approval. Present a restart approval table: who re-enters, at which step, with which (new) copy, at what daily cap. Wait for explicit approval.

7. Re-enroll safely

  • Run validate_flow_reenroll first for contacts being re-enrolled - never skip the validation step.
  • Re-enroll approved contacts with reenroll_flow_enrollments; enroll net-new approved contacts with enroll_awaiting_flow_items.
  • Republish the flow with manage_flow_publication.
  • Confirm with list_enrollments that re-enrolled contacts moved from awaiting to in-progress, and report the restart state to the user.

Duplicate-send rule: anyone who already received a step's message must never receive that step again. If FirstTouch cannot confirm where a contact stopped, leave that contact out and list them for manual review - skipping a prospect is recoverable; double-messaging them is not.

Output

  • Pre-pause snapshot (enrolled / sent / queued / awaiting)
  • Diagnosis: copy, audience, volume, or wrong-motion
  • Canceled-enrollment list with steps reached
  • Restart approval table and post-restart confirmation

Examples

User: "The AI SDR batch is getting bad replies - stop it." Play: pause via manage_flow_publication → snapshot shows 34 sent, 61 queued → diagnosis: opener has a meeting ask (iron-rule violation) → cancel the 61 queued enrollments, fix copy through the messaging quality gate → restart approval at 10/day → validate_flow_reenroll + reenroll_flow_enrollments → confirm in-progress. The 34 already-sent contacts are excluded from the restart.

Why this play wins

Every outbound motion eventually needs a mid-flight correction. Without a clean pause-and-fix path, users either let a broken campaign keep sending or nuke everything and double-touch people on the redo. This play is the difference between a five-minute correction and a burned account.

Pitfalls

  • Pausing without snapshotting first - you can't restart safely if you don't know who got what.
  • Canceling everyone by default - completed contacts don't need cancellation; untouched rows may only need the fix.
  • Treating the restart as pre-approved - the original approval covered the original copy/audience/caps. Fixed motion = fresh gate.
  • Re-enrolling without validate_flow_reenroll - the validation exists to catch exactly the duplicate-send cases humans miss.

Reference

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most marketing audience skills give in ~1.5k tokens

Counted across 690 of the 894 authors here whose files we hold, read 2026-08-07

  • Apply Poppins font to headingsin 41 of 690, across 6 files
  • Apply Lora font to body textin 41 of 690, across 6 files
  • Use Arial fallback for headingsin 39 of 690, across 4 files
  • Use Georgia fallback for body textin 39 of 690, across 4 files
  • Maintain text hierarchy and formattingin 39 of 690, across 4 files
  • Use accent colors for non-text shapesin 38 of 690, across 3 files
  • Use RGB values for precise color matchingin 38 of 690, across 3 files
  • Use brand colors for primary text and backgroundsin 36 of 690, across 1 file
  • Read product marketing context file before asking questions, starting, or auditingin 35 of 690, across 23 files
  • Use active voice instead of passive voicein 26 of 690, across 10 files
  • Implement or generate appropriate JSON-LD structured datain 24 of 690, across 17 files
  • Prioritize clarity over clevernessin 22 of 690, across 8 files

Said here and by no other author read

  • complete onboarding before first run
  • confirm the exact flow plan with the user
  • snapshot enrollment status before pausing
  • pause the flow to stop new sends
  • diagnose the actual problem type
  • scope cancellation to affected contacts only

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.