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.
In this repository23 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- ambiguity-resolverSkill
- challengeSkill
- deep-thinkSkill
- design-critique-templateSkill
- experiment-designSkill
- ingestSkill
- interview-analysisSkill
- multi-source-signal-synthesiserSkill
- next-advisorSkill
- pipeline-stepsSkill
- post-launch-reviewSkill
- problem-structuringSkill
- product-requirements-docSkill
- solution-scoringSkill
- strategic-narrative-generatorSkill
- system-design-docSkill
- technical-spec-documentSkill
- tracking-and-funnelsSkill
- ui-pattern-librarySkill
- user-persona-builderSkill
- user-story-generatorSkill
- user-testingSkill
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
- Distinctive: line 21
Before assembling a Gate presentation (gate preconditions require Frame)
- Distinctive: line 22
When the PM asks about targets / success criteria
- Very specific: line 23
When /next recommends it (Frame gaps + a gate is near)
- Distinctive: line 24
NEVER as a forced first step of a new initiative
- Distinctive: line 33
If PM doesn't know exact numbers: "Give your best estimate — we'll refine after step 1"
- Distinctive: line 40
If PM says "just me": note that, but ask about dev lead and designer
- Distinctive: line 45
If no OKR: flag as risk for report presentation ("needs strategic framing")
- Distinctive: line 60
If PM resists: "This protects your time.
- Distinctive: line 61
Default suggestion: "If step 6 validation confirms none of the hypotheses"
- Very specific: line 71
If rich data exists: suggest /ingest right after the checklist — it will move hypotheses to REAL fastest
- Very specific: line 77
If they say "I'll figure it out later" — set None, can be changed before /create-tickets
- Distinctive: line 94
Don't make this feel like a bureaucratic form.
- Distinctive: line 95
If PM has already filled some fields in CONTEXT.md — acknowledge and skip those.
- Distinctive: line 96
If PM wants to skip the checklist: "I understand, but 5 minutes now saves hours later.
≈ “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.
Files
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
/nextrecommends 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
/ingestright 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
- CONTEXT.md — all checklist answers written to appropriate fields
- 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.
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
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.

