agentleFS
megandmartinmegandmartin/agent-skills-repo/skills/business-ops/pipeline-tracker/SKILL.md

pipeline-tracker

Skill4 starsChanged 4 months agoMITView on GitHub

Maintain a lightweight CRM in a local leads.csv (stage, value, next_action, last_touch) using python3 stdlib, and report a pipeline summary on demand. Use when the user asks to "add a lead", "update the pipeline", "move X to negotiation", "what's my pipeline worth", "who needs follow-up", or wants a simple CRM without a SaaS tool. Don't use for researching a lead (lead-research-dossier) or billing a closed-won deal (invoice-runner).

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    After outreach (cold-outreach-sequencer) or a call, to log the touch and next action.

    line 21

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Pipeline Tracker

A zero-dependency CRM: one leads.csv in the project folder, edited safely with python3's stdlib csv module, plus an on-demand pipeline summary (totals by stage, overdue follow-ups). No SaaS, no lock-in — the file is the database.

When to Use

  • User wants to add, update, or move a lead through stages.
  • User asks "what's in my pipeline?", "total pipeline value?", or "who's gone cold?"
  • After outreach (cold-outreach-sequencer) or a call, to log the touch and next action.
  • Not for: researching a lead (lead-research-dossier) or billing a closed-won deal (invoice-runner).

Quick Reference

ActionCommand / Call
Init file (once)python3 -c "import csv,os; os.path.exists('leads.csv') or csv.writer(open('leads.csv','w',newline='')).writerow(['name','company','stage','value','next_action','last_touch'])"
View allpython3 -c "import csv; [print(r) for r in csv.reader(open('leads.csv'))]"
Add / update / summarypython3 heredoc scripts in the Procedure
Valid stagesnew → contacted → replied → call → proposal → negotiation → won | lost
Backup before writecp leads.csv leads.csv.bak

Procedure

  1. Precheck — command -v python3; locate leads.csv in the working directory (ask the user where it lives if not found). If absent, run the init one-liner from Quick Reference. Success: file exists with the 6-column header.
  2. Backup — before any write, cp leads.csv leads.csv.bak. Cheap insurance.
  3. Add a lead — run:
    python3 - "$NAME" "$COMPANY" "$VALUE" "$NEXT" <<'EOF'
    import csv, sys, datetime
    name, company, value, nxt = sys.argv[1:5]
    rows = list(csv.reader(open('leads.csv')))
    if any(r and r[0].lower()==name.lower() and r[1].lower()==company.lower() for r in rows[1:]):
        sys.exit(f"DUPLICATE: {name} @ {company} already tracked")
    rows.append([name, company, 'new', value, nxt, datetime.date.today().isoformat()])
    csv.writer(open('leads.csv','w',newline='')).writerows(rows)
    print(f"Added {name} @ {company} (stage=new, value={value})")
    EOF
    Success: "Added ..." printed; duplicates rejected.
  4. Update a lead (stage move, new next_action, touch logged) — same pattern: load all rows, match on name+company (case-insensitive), set stage/value/next_action, always refresh last_touch to today, rewrite the whole file with newline=''. Reject stage values outside the Quick Reference list. Success: script prints the before → after line.
  5. Summary on demand — run:
    python3 <<'EOF'
    import csv, datetime, collections
    rows = list(csv.DictReader(open('leads.csv')))
    by = collections.defaultdict(lambda: [0, 0.0])
    for r in rows:
        by[r['stage']][0] += 1
        try: by[r['stage']][1] += float(r['value'] or 0)
        except ValueError: print(f"WARN bad value: {r['name']}: {r['value']!r}")
    openv = sum(v for s,(c,v) in by.items() if s not in ('won','lost'))
    print(f"Pipeline: {len(rows)} leads | open value ${openv:,.0f}")
    for s,(c,v) in sorted(by.items()): print(f"  {s:12} {c:3}  ${v:,.0f}")
    today = datetime.date.today()
    stale = [r for r in rows if r['stage'] not in ('won','lost')
             and (today - datetime.date.fromisoformat(r['last_touch'])).days > 7]
    print("Needs follow-up (7+ days):")
    for r in stale: print(f"  {r['name']} @ {r['company']} — {r['next_action']} (last: {r['last_touch']})")
    EOF
  6. Deliver — present the summary using the Output Template; on writes, confirm what changed and that the backup exists.

Output Template

## Pipeline Summary — <date>
Total leads: N | Open pipeline value: $X
| Stage | Count | Value |
|---|---|---|
### Needs follow-up (no touch in 7+ days)
- <name> @ <company> — next: <next_action> — last touch <date>
### Changes this session
- <added/updated rows, before → after>

Pitfalls

  • Commas or quotes inside fields corrupt the file — hand-editing "Acme, Inc." with echo >> breaks the column count. Recovery: always write via the csv module (it quotes correctly); if the file is already broken, find the bad row with a len(row) != 6 check and repair from leads.csv.bak.
  • Duplicate leads inflate the pipeline — the same prospect added twice doubles their value in totals. Recovery: the add script's name+company check blocks this; if dupes exist, keep the row with the later last_touch and delete the other.
  • Non-numeric value breaks the summary — "$5k" or "TBD" won't float(). Recovery: the summary warns instead of crashing; store plain numbers (5000, not $5k) and fix flagged rows immediately.
  • Invalid stage names fragment reporting — "Negotiating" vs "negotiation" become two buckets. Recovery: validate against the canonical stage list on every write; normalize to lowercase before saving.

Verification

  • leads.csv still parses: every row has exactly 6 columns
  • All stages in file are from the canonical list; all values are numeric
  • last_touch updated to today on every row touched this session
  • leads.csv.bak exists from before this session's writes
  • Summary totals match a manual spot-check of 2–3 rows

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.