agentleFS
link7373link7373/agentic-sales-marketing/.claude/skills/review-monthly/SKILL.md

review-monthly

Skill1 starsChanged 3 months agoMITView on GitHub

The monthly retro — strategy versus actuals, budget and ROI close-out, competitor refresh, decision-log audit, learning consolidation, and proposed strategy amendments. Run on the first Monday of each month; it's the team's defense against silent drift.

The file

Files
SKILL.md

/review-monthly — Zoom Out Before Zooming In

Inputs: the last 4 weekly scorecards in knowledge/results-history.md, knowledge/budget.md, knowledge/strategy.md, knowledge/decision-log.md, knowledge/competitors.md, the What we've learned here sections across knowledge/methods/. Agents: Head of Growth leads; growth-analyst supplies the numbers; market-researcher refreshes competitive intel; the relevant strategist joins if an amendment touches their domain. Method: knowledge/methods/funnel-and-measurement.md (reporting rhythm), knowledge/methods/digital-channel-strategy.md (quarterly/monthly review loop).

Weekly /measure asks "did this week work?" This asks the bigger question: is the plan still the right plan? A team that only ever runs weekly reviews optimizes its way down a dead end very efficiently.

Steps

  1. Assemble the month. Four scorecards into one view: funnel numbers by month, not week (weekly noise disappears at monthly scale). Sales, cost per sale, and the trend direction on each stage of the funnel.
  2. Close the books. From knowledge/budget.md: approved vs spent vs return per channel for the month. Which channel earned its money, which didn't, which we can't tell yet — and honestly which. Append the month's close-out to the budget history.
  3. Judge strategy against actuals. For each channel in knowledge/strategy.md: is it hitting the targets set at approval time? Has any channel passed its pre-committed pivot trigger? Is the first-sale (or growth) milestone still on the dated path, and if not, by how much?
  4. Refresh the competitive picture. market-researcher updates knowledge/competitors.md: pricing changes, new entrants, positioning shifts, review-sentiment moves. Anything here that threatens our positioning gets flagged to brand-strategist.
  5. Audit the decision log. Walk knowledge/decision-log.md for the trailing quarter. Flag: decisions the results now contradict, decisions made on assumptions since tested, and decisions everyone has quietly stopped following (drift caught in the act). Each gets kept, superseded with a new dated entry, or explicitly retired.
  6. Consolidate the learnings. Read every method file's What we've learned here. Promote patterns that appeared three or more times into the strategy or the relevant knowledge file (offer, positioning, channel plan) — a repeated learning is no longer a learning, it's a rule. Prune lines that later evidence contradicted, with a note rather than a deletion.
  7. Propose amendments. Write the month's recommendation: keep / adjust / pivot per channel, budget reallocation, and any offer or positioning change the evidence demands. Each with the number that argues for it. Operator approval of strategy amendments is the anti-drift gate — the strategy changes deliberately and on the record, or it doesn't change.
  8. Set next month. Targets, the approval batch to queue, and the one thing the team will do differently. Confirm the scheduled tasks are still alive (else run /health-check).

Before you start

  1. Required inputs: knowledge/strategy.md approved; at least two weekly scorecards in knowledge/results-history.md — with less history, run /measure and say so rather than manufacturing a trend.
  2. Scan knowledge/decision-log.md in full (this skill audits it — reading it is step 5).
  3. Gate check: proposing amendments needs no approval; adopting them does. Nothing in knowledge/strategy.md changes without the operator's yes, logged.

If things go wrong

  • Data is missing for part of the month → report the gap, judge on what exists, and route the instrumentation fix to /connect-tools.
  • Results argue for a pivot the operator resists → present the numbers once, plainly, with the cost of waiting another month; then respect their call and log it as a decision with its rationale. It's their business.
  • Operator unavailable → park the amendment proposal in knowledge/pending.md; the current strategy stays in force (drift is what happens when the team acts on an unapproved amendment).

Outputs

The monthly review on templates/monthly-review.md, appended to knowledge/results-history.md; refreshed knowledge/competitors.md; an audited decision log; consolidated learnings; the amendment proposal. Close with the report per templates/skill-report.md.

Writeback

knowledge/results-history.md, knowledge/budget.md (month close-out), knowledge/competitors.md, knowledge/decision-log.md (superseding and retiring entries), knowledge/methods/*.md (pruned/promoted learnings), knowledge/strategy.md (only after approval), approval to knowledge/approvals-log.md.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.