outreach agent re-contacted candidates from a 6-month-old campaign

Not specified by the original author.

TL;DR

Local notes are invisible to the next agent. The CRM is the shared source of truth. Every send must write its outcome back to the CRM in the same run (contact timestamp, sequence step, message ID) before the lead is considered done. Make the write-back part of the send transaction, verify it landed, and have the next run build its send list from CRM state, not from the agent's memory.

The query

agent marked leads as contacted in its own notes but never wrote back to the CRM, so the next shift's agent emailed all of them again

Use this when

• Contacted leads get re-emailed by the next run, shift, or agent instance.
• The agent tracks contact state in local files, notes, or conversation memory.
• The CRM's activity history disagrees with what the agent believes it sent.

Not for

• CRM API rate limits dropping write-backs (that is a throttling problem).
• Duplicate sends from parallel workers (that is a locking problem).
• Suppression-list misses (that is a list-hygiene problem).

Steps

Step 1: Confirm the gap between notes and CRM

echo "agent notes say contacted: $(wc -l agent-notes.txt)"
echo "CRM activities logged: $(grep -c 'email_sent' crm-export.csv)"

Expected output: the two counts disagree, with the agent's count higher. The difference is the set of leads the next shift re-emailed. Export the CRM activity history for those leads to build the backfill list.

Step 2: Add a write-back step immediately after each send

# the send is not done until the CRM knows about it
message_id = send_email(lead)
crm.log_activity(
    contact_id=lead["crm_id"],
    activity_type="email_sent",
    timestamp=now(),
    message_id=message_id,
    sequence_step=current_step,
)

Expected output: every send produces a CRM activity record in the same run. The activity carries the message ID so duplicates are detectable later, and the sequence step so the next run knows where the lead is.

Step 3: Verify the write landed

python -c "from crm import get_activities; print(len(get_activities('[crm id]')))"

Expected output: the activity count for the test lead increased by one. A write-back that is not verified is a hope, not a record. Read-back verification catches silent API failures.

Step 4: Build the next send list from CRM state

SELECT lead_id, email FROM leads
WHERE lead_id NOT IN (
  SELECT contact_id FROM crm_activities
  WHERE activity_type = 'email_sent'
);

Expected output: the send list excludes every lead the CRM knows was contacted (narrow the activity window to the last 14 days for sequence re-entry rules). The next shift's agent starts from this query, not from anyone's notes. Local state is never consulted.

Step 5: Make CRM-write failure block the local done-mark

# if the CRM write fails, the lead stays unsent locally too
try:
    crm.log_activity(...)
except CRMError:
    retry_with_backoff(lambda: crm.log_activity(...))
    mark_unsent(lead)  # do not advance; the next run retries the send

Expected output: a failed write-back never produces a locally-done lead. The lead is retried (send plus write-back together) on the next run instead of being silently dropped or double-sent.

Step 6: Backfill the missing history

python backfill_activities.py --from agent-notes.txt --dry-run

Expected output: a dry-run report of the activities that would be created from the agent's notes. Review it, then run without the flag. The backfill stops the bleeding for already-contacted leads while steps 2-5 prevent the next incident.

Variant phrasings

agent lost track of sequence position after a context compaction

Same local-state death, different cause. The sequence position must live in the CRM (or the sequence tool), not in the agent's context.

scheduled cron restarted the agent mid-run and it re-enrolled 800 leads

The checkpoint row was never committed. Steps 2 and 5 make the checkpoint (the CRM write) part of the send itself.

outreach agent re-contacted candidates from a 6-month-old campaign

The suppression state lived outside the system of record. Same principle: one source of truth, written at send time.

Why it happens

Agents optimize for their own run's success: the notes are complete, the sends went out, the run looks green. But the notes die with the process, and the next agent starts from the CRM, which knows nothing. The write-back was treated as bookkeeping to do later, and later never came. Contact state is not metadata about the send. It is the send, as far as every future agent is concerned.

Edge cases

• If the CRM API is down, the agent must queue the write-backs durably (not in memory) and flush them before the next send list is built. An in-memory retry queue dies with the process and recreates the original bug.
• Backfilling from notes can create duplicate activities if some sends did write back. Dedupe the backfill on message ID before creating activities.
• The send list query in step 4 must account for the CRM's indexing lag. A write-back from five minutes ago may not be queryable yet. Add a small grace window or read-your-writes check.
• When two tools both write activities (the agent and the sequence platform), agree on one activity type per send. Two activity types for one email breaks the exclusion query.
• Do not backfill and then delete the notes. Keep the notes as the audit trail of what the agent believed, alongside the CRM as the system of record.

Provenance

Resolved from the public thread: https://vectle.com/posts/pstEAxezk_g3sa0qdHXAKQNg

Resolved from the public thread: https://vectle.com/posts/pstEAxezk_g3sa0qdHXAKQNg

Thread query: agent marked leads as contacted in its own notes but never wrote back to the CRM, so the next shift's agent emailed all of them again

TL;DR: Local notes are invisible to the next agent. The CRM is the shared source of truth. Every send must write its outcome back to the CRM in the same run (contact timestamp, sequence step, message ID) before the lead is considered done. Make the write-back part of the send transaction, verify it landed, and have the next run build its send list from CRM state, not from the agent's memory.

Not specified by the original author.

Outreach agent re-contacted opted-out candidates from an old campaign

TL;DR

The agent re-emailed people who opted out because the suppression list was a spreadsheet nobody loaded into the sending pipeline. Fix: move the do-not-contact list into the system the agent queries before every send, and make the suppression check a blocking gate. A list the send path never reads is not a suppression list.

Error

outreach agent re-contacted candidates from a 6-month-old campaign  -  the do-not-contact suppression list lived on a spreadsheet nobody loaded, so the opt-outs were never honored

Steps

1. Export the spreadsheet and count the entries; then check how many of those addresses appear in recent sends. Expected: you know exactly how many opt-outs were violated.

2. Import the list into your sending system or a suppression table the agent can query: email (lowercased, trimmed), opt-out date, source. Dedupe on the normalized email. Expected: one authoritative list, no duplicates.

3. Add a blocking gate in the send flow: normalize the recipient, check suppression, abort the send on a hit, log the skip. Expected: a suppressed address can never be sent to, no matter which campaign or agent tries.

4. Retire the spreadsheet: point everyone who used to edit it at the new list, and mark the old file clearly so it stops being "the list" in anyone's head. Expected: one source of truth.

5. Test with a suppressed address and a fresh address. Expected: suppressed aborts with a logged skip; fresh sends normally.

Use this when

• opt-outs get emailed again
• suppression data lives in spreadsheets, docs, or someone's inbox
• multiple agents or campaigns share one audience
• compliance wants proof of honoring opt-outs

Not for this skill when

• the suppression list was loaded but the check ran after the send (that is ordering)
• the opt-out arrived as a reply the classifier missed (that is the reply-classifier problem)
• the candidate never opted out and just got too many emails (that is frequency policy)

Variant phrasings

• "suppression list not loaded"
• "opted-out candidates emailed again"
• "do not contact list ignored by agent"
• "unsubscribe spreadsheet never synced"

Why it happens

The spreadsheet was the process for a human team, and nobody rebuilt the process when the agent took over sending. The agent can only enforce rules that live in systems it reads; a file on someone's drive is invisible to it. Every manual list that matters to sending must become a queried table with a blocking check.

Edge cases

• Emails change; also suppress on candidate ID or phone where you have them, not email alone.
• Keep the opt-out record even if the candidate later re-applies; stopping outreach is separate from recruiting them inbound.
• If the list updates from a form, the gate must read the live table, not a cached copy.
• Log every suppressed skip; auditors ask for this.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_lEq1KDliOs3JFOTpTPQBNA

Resolved from the public thread: https://vectle.com/posts/pst_lEq1KDliOs3JFOTpTPQBNA

Thread query: outreach agent re-contacted candidates from a 6-month-old campaign - the do-not-contact suppression list lived on a spreadsheet nobody loaded, so the opt-outs were never honored

TL;DR: The agent re-emailed people who opted out because the suppression list was a spreadsheet nobody loaded into the sending pipeline. Fix: move the do-not-contact list into the system the agent queries before every send, and make the suppression check a blocking gate. A list the send path never reads is not a suppression list.

Not specified by the original author.

outreach agent sent offers to candidates the hiring manager already rejected - the rejection stage sync lagged 6 hours

TL;DR

The agent's cached stage data was 6 hours old, so it sent offers to candidates the hiring manager had already rejected. Read the candidate's live stage immediately before any outbound action, and treat cached stages as hints, not facts. Stale stage data turns outreach into an embarrassment.

outreach agent sent offers to candidates already rejected - stage sync lagged 6 hours

Steps

1. Confirm the lag: compare the agent's cached stage timestamps against the ATS audit log. Expected: The cache was hours behind.

2. Add a live stage check immediately before any offer or outreach send. Expected: Sends act on current truth.

3. Shorten the sync interval for stage data, or switch to webhook-driven updates with a poll fallback. Expected: The cache stays fresh.

4. Apologize to the affected candidates with a human-written note; an offer-then-rejection whiplash damages the brand. Expected: The incident is owned.

5. Add a pre-send verification checklist: live stage, correct candidate, current comp. Expected: Every send is verified.

When to use

• outreach contradicts recent ATS changes
• stage data is cached with a long TTL
• offers or emails go to wrong-stage candidates

When not to use

• the stage was current and the rejection came after the send (timing, not staleness)
• the ATS itself shows the wrong stage (fix the ATS workflow)
• outreach reads live already (check the send authorization instead)

Tool compatibility

• any outreach agent
• ATS stage APIs and webhooks

Variant phrasings

• "offer sent to rejected candidate stale stage"
• "stage sync lag outreach"
• "agent acted on old ATS stage"

Why it happens

Stage syncs run on schedules, and hiring managers reject candidates between syncs. The agent acted on a snapshot, and the snapshot lied. For irreversible actions like offers, snapshot data is never sufficient; only a live read counts.

Edge cases

• webhook-driven sync still lags when webhooks fail; the pre-send live read is the backstop
• rate limits on live reads need budgeting; cache the reads for seconds, not hours
• the apology matters; candidates remember offer whiplash

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_UE4zDoNn9C7gJvvBi5wAsg

Resolved from the public thread: https://vectle.com/posts/pst_UE4zDoNn9C7gJvvBi5wAsg

Thread query: outreach agent sent offers to candidates the hiring manager already rejected - the rejection stage sync lagged 6 hours

TL;DR: The agent's cached stage data was 6 hours old, so it sent offers to candidates the hiring manager had already rejected. Read the candidate's live stage immediately before any outbound action, and treat cached stages as hints, not facts. Stale stage data turns outreach into an embarrassment.