LeadPanther API

LinkedIn Safety

LeadPanther can schedule LinkedIn-facing work such as lead magnet follow-up DMs, follow-up public comments, workspace comments, and workspace reposts. These actions should be scheduled like real human activity, not like a batch job.

The rule is simple: never schedule a large group of LinkedIn actions at the same timestamp or at exact repeating intervals. Randomize timing, spread actions across working hours, and keep total action volume appropriate for the account.

What This Applies To

This page applies to LeadPanther actions that cause LinkedIn activity:

  • lead magnet follow-up DMs after a resource DM is delivered
  • lead magnet follow-up public comments after a public comment reply is delivered
  • workspace social action comments
  • workspace social action reposts
  • future LinkedIn-backed write actions added to API v1

This page does not apply to ordinary LeadPanther API read calls that only read LeadPanther data, such as listing posts, listing leads, reading activity, or reading analytics.

Required Scheduling Pattern

Do this:

  • Randomize each action's due time.
  • Spread actions across normal working hours for the LinkedIn account.
  • Use multiple small launches instead of one large burst.
  • Start lower for new, inactive, or low-trust accounts, then increase gradually.
  • Keep skip_if_replied enabled for lead magnet follow-ups unless there is a clear reason to override it.
  • Cancel pending actions when the recipient has already replied or the context is no longer relevant.
  • Monitor failed, skipped, rate-limited, and provider-error outcomes before increasing volume.

Do not do this:

  • Do not make every follow-up due at the same delay_minutes.
  • Do not schedule every workspace comment exactly 1, 2, or 5 minutes after publish.
  • Do not queue hundreds of comments, DMs, or reposts for the same LinkedIn account at one time.
  • Do not run the same schedule every day at the same clock time.
  • Do not retry ambiguous provider sends blindly.
  • Do not use fake, new, or inactive LinkedIn accounts for high-volume activity.

Daily Planning Range

For established, active LinkedIn accounts with healthy history, LeadPanther's planning range is roughly 500 to 1,000 LinkedIn provider actions per account per day across relevant action types. This is not a guarantee, entitlement, or public API quota. Account age, account trust, profile history, recipient behavior, message quality, reply rates, and LinkedIn-side changes can all lower the safe amount.

For newer, inactive, or lower-trust accounts, start far below that range and ramp gradually. A safe scheduler should treat provider warnings, repeated failures, suspicious verification prompts, or sudden drops in delivery as a signal to slow down or stop.

Lead Magnet Follow-Ups

Lead magnet follow-up rules use delay_minutes. That delay is measured from the successful delivery event that materializes the follow-up action.

Good pattern:

  • First follow-up DM: randomized within a broad window after resource delivery.
  • Second follow-up DM: randomized on a later day or later working-hour window.
  • Public comment follow-up: randomized separately from DM follow-ups.
  • Different campaigns use different timing ranges.

Bad pattern:

  • Every recipient gets delay_minutes: 60.
  • Every campaign uses the same 24-hour and 72-hour offsets.
  • A large import causes hundreds of actions to become due at the same minute.

LeadPanther's worker includes safeguards such as one action per user per run, same-user cooldown, and short random waits between processed sends. Those safeguards reduce burst risk but do not make unsafe scheduling safe. If many actions become due at once, the backlog can still create an obvious automation pattern.

Workspace Comments And Reposts

Workspace social actions use scheduled_for. The API stores the timestamp you send. It does not rewrite your schedule into a safer pattern.

For workspace comments:

  • Avoid scheduling all comments immediately after publish.
  • Use a randomized window after the post is live.
  • Avoid having every workspace member comment on every post.
  • Vary actor selection, timing, and comment content.

For workspace reposts:

  • Schedule reposts later than comments.
  • Randomize repost timing across a wider window.
  • Avoid repeated same-actor/same-delay patterns.

Use idempotency keys to make retries safe, but do not use retries to push duplicate or near-duplicate actions.

Connection Requests

LeadPanther API v1 does not schedule outbound LinkedIn connection requests. Do not build a connection-request workflow by guessing an unsupported endpoint.

Inbound connection acceptance is not a LeadPanther outbound follow-up action. If LeadPanther later adds outbound connection request support, it needs a separate product and security review because invitation limits and account risk are different from inbound lead magnet follow-ups.

Relevant And Non-Relevant Provider Guidance

Relevant to LeadPanther:

  • LinkedIn account trust matters.
  • Real, active accounts are safer than new or fake accounts.
  • Random timing is safer than fixed timing.
  • Working-hours distribution is safer than a single daily burst.
  • Provider errors and warnings should slow or stop automation.
  • Comments, DMs, reposts, profile retrieval, and company/post retrieval should be minimized and spaced.

Not relevant to current LeadPanther API v1:

  • Non-LinkedIn provider limits from third-party provider docs.
  • Recruiting-only LinkedIn product behavior.
  • Paid-message quotas outside LeadPanther's current API surface.
  • Outbound connection request automation.
  • Any route not listed in the LeadPanther API reference.

Operational Checklist

Before enabling a high-volume workflow:

  1. Confirm the account is active, real, and healthy.
  2. Confirm the workflow only uses documented LeadPanther API routes.
  3. Generate randomized due times before calling the API.
  4. Spread actions across working hours.
  5. Keep action volume within the account's current risk tolerance.
  6. Review pending actions before they execute.
  7. Monitor skipped, failed, provider-error, and rate-limited outcomes.
  8. Pause or cancel pending actions when provider risk rises.

If you are unsure whether a schedule is safe, slow it down and spread it out. A delayed follow-up is better than a banned or restricted LinkedIn account.