# Cold Email Weekly Rhythm

> Operational cadence playbook for the cold-email program. Defines what happens on Monday (PSP signals + list refresh), Wednesday (sequence ship + reply triage), Friday (reply scoring + adjust). Surfaces this week's queue, drift signals, and the metric review against last week. Inspired by coldoutboundskills' /cold-email-weekly-rhythm pattern — the cadence is what separates hobbyist programs from compounding ones.

- Skill: `cmj-hub/cold-email-weekly-rhythm` (Agent Skill)
- Install (CLI): `npx skillmds@latest add cmj-hub/cold-email-weekly-rhythm`
- Raw SKILL.md: https://api.skillmd.com/api/skills/cmj-hub/cold-email-weekly-rhythm/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Productivity
- License: MIT
- Author: cmj-hub (https://skillmd.com/u/cmj-hub)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/cmj-hub/cold-email-weekly-rhythm

---


# Cold Email Weekly Rhythm — operational cadence

The framework + the infrastructure are necessary but not sufficient.
What separates programs that ship from programs that don't is the
**weekly cadence**. This sub-skill walks the operator through that
cadence and surfaces the queue for this week.

## Activation

Loaded by `cold-email-kickoff` when the operator is in iteration mode
(reply data available, brand-config set up).

User-invocable on demand:
- "What's my cold-email rhythm this week"
- "Run the weekly rhythm"
- "Monday queue"
- "Friday review"

## The default rhythm (Mon / Wed / Fri)

### Monday — PSP signals + list refresh

Goals:
- Refresh the list with this week's signals (≤14 days old)
- Score list quality (dedup, role-fit, freshness)
- Set the week's send target (volume × personalization depth)

Tasks the skill runs:

1. **Pull signals**: For each anchor in `brand-config.psp.signal_anchors`,
   surface 5-10 new candidates from the operator's sources (LinkedIn job
   search, Crunchbase, RSS, etc.)
2. **Score the list**: Invoke `cold-email-list-quality` for dedup +
   role-fit + freshness scoring
3. **Set the queue**: Recommend send volume based on:
   - Warm-up state (`brand-config.infrastructure.warm_up_status`)
   - Reply-rate trend from last week
   - Available personalization time

Output:

```markdown
# Monday — <date>

## This week's send queue
- Volume: <N> sends across <M> sequences
- Personalization depth: deep (≤30/seq) / medium (≤80/seq) / light (≤200/seq)

## Signal hunt (top 5 to verify by Wednesday)
1. <Company> — <signal>
2. <Company> — <signal>
3. ...

## List quality
- Score: <0-100>
- Dedup: <N> removed
- Role-fit: <N> outside ICP — flagged for removal
- Freshness: <N> stale signals removed

## Next: Wednesday ship + Friday review
```

### Wednesday — sequence ship + reply triage

Goals:
- Ship this week's sequence (T1 for new prospects, T2/T3/T4 for in-flight)
- Triage reply pile from earlier in the week

Tasks:

1. **Draft sequences** via `cold-email-craft` using the operator's
   brand-config + SOUL.md voice
2. **Self-check each draft** via `cold-email-reviewer` agent
3. **Spam-lint** via `cold-email-spam-lint`
4. **Route to send infrastructure** (the operator's actual sending stack)
5. **Triage replies** received since Monday:
   - Score each via `cold-email-reply-scoring`
   - Route per `brand-config.operations.reply_routing`

Output:

```markdown
# Wednesday — <date>

## Sequences shipping today
- T1 (new): <N> drafts, <N> passed self-check
- T2/T3/T4 (in-flight): <N> sends

## Reply triage (since Monday)
- Buy-signal: <N> (route to sales)
- Positive: <N> (reply within 4 hours per ops)
- Neutral: <N> (add to nurture stream)
- Not-interested: <N> (suppress + log reason)

## Flagged for review
- <draft> failed self-check on <reason>
- <reply> ambiguous — needs human read
```

### Friday — score + adjust

Goals:
- Compute this week's metrics
- Compare to last week
- Surface 1 adjustment for next week (just one — don't compound changes)

Tasks:

1. **Compute weekly metrics** from logs:
   - Sends, opens (if tracked), replies, positive-reply rate, buy-signal rate
   - Time-to-reply distribution
   - Top-performing subject lines / first-line patterns
2. **Compare to last week** + last 4-week trend
3. **Surface 1 lever to adjust** next week — the highest-leverage
   single change

Output:

```markdown
# Friday — <date>

## This week's numbers
| Metric | This week | Last week | 4-wk trend |
|---|---|---|---|
| Sends | <N> | <N> | ↑ / ↓ / → |
| Reply rate | <%> | <%> | ↑ / ↓ / → |
| Positive-reply rate | <%> | <%> | ↑ / ↓ / → |
| Buy-signal rate | <%> | <%> | ↑ / ↓ / → |

## What worked
- Top subject pattern: "<pattern>" — <N>% open
- Top first-line: signal-quoted opener with <pattern>

## What didn't
- <observation with data>

## Single lever for next week
<ONE specific change. Not a list. One.>
Reason: <one sentence>
Verify by: <metric to watch>
```

## Friday review — quarterly meta

Every 13 weeks (quarter), the skill also runs a meta-review:

- ICP precision check — has it tightened or drifted?
- PSP refresh — have signals decayed? Is the vocabulary still current?
- EVP review — is the line still landing, or has the market caught up?
- Infrastructure age check — is the sending domain still in good shape?
- Banned-phrases list — any new patterns the operator should add?

The meta-review surfaces a single Q-level adjustment, not weekly-level
tweaks.

## Why cadence beats tactics

The framework is 80% of the lever. The remaining 20% is whether the
operator runs it weekly without skipping. Skipping a week is fine.
Skipping two compounds — the list goes stale, signals decay, replies
get cold, the sender's confidence drops.

The skill enforces the cadence by surfacing what's due TODAY when
invoked. No project-management software, no big planning ritual —
just "here's what Monday needs from you."

## References

- `../cold-email-list-quality/SKILL.md` — Monday list scoring
- `../cold-email-craft/SKILL.md` — Wednesday drafting
- `../cold-email-reply-scoring/SKILL.md` — Wednesday + Friday triage
- `../cold-email-spam-lint/SKILL.md` — Wednesday pre-send gate
- `../cold-email-audit/SKILL.md` — quarterly meta-review
- `../../scripts/` — Python scoring scripts the skills shell out to