agentleFS
lenar-amirovlenar-amirov/product-pipeline-public/.claude/skills/setup-initiative/SKILL.md

setup-initiative

Skill13 starsChanged 5 months agoMITView on GitHub

Guides the PM through an initiative alignment checklist — metric, baseline, target, stakeholders, success/kill criteria. Use before a gate, when the PM says "setup initiative", "set the target", "define success criteria", "align on goals", or when the coverage map shows Frame incomplete. NOT a prerequisite for starting work — jobs run without it.

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 11 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before assembling a Gate presentation (gate preconditions require Frame)

    line 21
  2. Distinctive:

    When the PM asks about targets / success criteria

    line 22
  3. Very specific:

    When /next recommends it (Frame gaps + a gate is near)

    line 23
  4. Distinctive:

    NEVER as a forced first step of a new initiative

    line 24
  5. Distinctive:

    If PM doesn't know exact numbers: "Give your best estimate — we'll refine after step 1"

    line 33
  6. Distinctive:

    If PM says "just me": note that, but ask about dev lead and designer

    line 40
  7. Distinctive:

    If no OKR: flag as risk for report presentation ("needs strategic framing")

    line 45
  8. Distinctive:

    If PM resists: "This protects your time.

    line 60
  9. Distinctive:

    Default suggestion: "If step 6 validation confirms none of the hypotheses"

    line 61
  10. Very specific:

    If rich data exists: suggest /ingest right after the checklist — it will move hypotheses to REAL fastest

    line 71
  11. Very specific:

    If they say "I'll figure it out later" — set None, can be changed before /create-tickets

    line 77
  12. Distinctive:

    Don't make this feel like a bureaucratic form.

    line 94
  13. Distinctive:

    If PM has already filled some fields in CONTEXT.md — acknowledge and skip those.

    line 95
  14. Distinctive:

    If PM wants to skip the checklist: "I understand, but 5 minutes now saves hours later.

    line 96

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “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

Setup Initiative

Fills the Frame phase of the coverage map. NOT a gatekeeper for starting work — jobs run without it (zero-setup principle); this skill earns its place before the first gate or whenever the PM asks about targets.

Purpose

Explicit alignment on goals, constraints, stakeholders, and success criteria. The coverage map (coverage.py) shows Frame N/4 until metric, baseline, target and kill criteria are recorded — gates are blocked while Frame is incomplete.

When to use

  • Before assembling a Gate presentation (gate preconditions require Frame)
  • When the PM asks about targets / success criteria
  • When /next recommends it (Frame gaps + a gate is near)
  • NEVER as a forced first step of a new initiative

Checklist

Guide the PM through these 10 areas. For each area, ask a focused question, then write the answer to CONTEXT.md.

1. Outcome

Ask: "What metric are we trying to improve, and by how much?"

  • Get: metric name, current baseline, target, measurement horizon
  • If PM doesn't know exact numbers: "Give your best estimate — we'll refine after step 1"
  • Required: at minimum the metric name and direction (up/down)

2. Stakeholders

Ask: "Who needs to approve this at Report presentation, and who might block it?"

  • Get: decision-maker, influencers, potential blockers
  • Map: name -> role -> concern/interest
  • If PM says "just me": note that, but ask about dev lead and designer

3. OKR Alignment

Ask: "Which company or team OKR does this serve?"

  • Get: specific OKR text or "exploratory / not tied to OKR"
  • If no OKR: flag as risk for report presentation ("needs strategic framing")

4. Constraints

Ask: "What can't we change, and what's our timeline?"

  • Get: timeline, budget, team capacity, tech limitations, political constraints
  • Distinguish hard constraints (deadline) from soft (preference)

5. Success Criteria

Ask: "If this initiative goes perfectly, what does the world look like in 3 months?"

  • Get: specific, measurable outcome
  • Push back on vague answers: "more users" -> "X% increase in Y metric"

6. Kill Criteria

Ask: "Under what conditions should we stop this initiative?"

  • Get: specific threshold (e.g., "if validation shows <5% of users affected")
  • If PM resists: "This protects your time. What would make you say 'this isn't worth it'?"
  • Default suggestion: "If step 6 validation confirms none of the hypotheses"

7. User Segment

Ask: "Who specifically are we building this for?"

  • Get: segment definition, size, platform, key behavior
  • Push for specificity: "all users" -> "users who [specific behavior]"

8. Available Data

Ask: "What do we already know? Analytics, research, customer feedback?"

  • Get: inventory of existing data sources
  • If rich data exists: suggest /ingest right after the checklist — it will move hypotheses to REAL fastest

9. Tracker

Ask: "Where do you want dev tickets to land — Jira, Linear, GitHub Issues, or none?"

  • Get: tracker system, project key/board name, standard labels
  • If they say "I'll figure it out later" — set None, can be changed before /create-tickets
  • Mention: "If you want me to push tickets directly, you'll need to connect the MCP — see Tracker integration section in README.md"

(Pipeline templates are deprecated — do not offer them. The PM runs the jobs they need; the coverage map shows what evidence is still missing. If the PM explicitly wants to disable an Optional/Recommended step, toggle it in pipeline_config.steps with a one-line warning about what it costs.)

Output

  1. CONTEXT.md — all checklist answers written to appropriate fields
  2. Re-run the coverage check (python3 tools/scripts/coverage.py <dir>) and show the PM the Frame line before/after — the visible payoff of 5 minutes of alignment

Tips

  • Don't make this feel like a bureaucratic form. It's a conversation.
  • If PM has already filled some fields in CONTEXT.md — acknowledge and skip those.
  • If PM wants to skip the checklist: "I understand, but 5 minutes now saves hours later. Let me ask the 3 most critical questions: metric, segment, and stakeholders."
  • Adapt language to PM's style — if they're brief, be brief. If they elaborate, explore.

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.