agentleFS
BagelHoleBagelHole/BagelSkills/skills/scope-sniper/SKILL.md

scope-sniper

Skill2 starsChanged 4 months agoMITView on GitHub

Use this skill when an ambitious product or engineering proposal must be reduced to the smallest credible proof of value. Identify the riskiest assumption, remove work that does not test it, define a thin end-to-end slice, and record explicit non-goals and expansion triggers so “MVP” does not become a disguised full product.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    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. Distinctive:

    If a required input is retrievable from available tools or artifacts, retrieve it rather than asking the user.

    line 41
  2. Distinctive:

    Complete when: every retained item contributes to proof or mandatory trust

    line 67
  3. Distinctive:

    Do not defer the measurement required to know whether the proof worked.

    line 86
  4. Distinctive:

    Do not call table stakes “nice to have” when their absence invalidates trust or evidence.

    line 87
  5. Very specific:

    Keep/cut/defer/manual decisions with rationale

    line 96

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Scope Sniper

Outcome

Cut scope until the next build proves something important rather than merely producing more software.

This skill changes the agent's process: it requires evidence, explicit claim strength, and a checkable completion boundary. Load references/playbook.md only when deeper prompts or rubrics are needed. Use assets/output-template.md when producing the final artifact.

When to use

  • The user asks for an MVP, v1, prototype, pilot, proof of concept, or smaller scope.
  • A plan contains many features, integrations, roles, or infrastructure choices before value is proven.
  • A team needs to meet a deadline without destroying the core user promise.

Do not use when

  • The work is mandated compliance or migration scope that cannot be traded away.
  • The product is already proven and the task is execution sequencing rather than scope discovery.

Required inputs

  • Target user and promised outcome
  • What must be learned or proven
  • Deadline, resources, and non-negotiable constraints
  • Current evidence and reusable assets

If a required input is retrievable from available tools or artifacts, retrieve it rather than asking the user. Ask only when a missing answer represents authority, preference, or risk that tools cannot resolve.

Workflow

1. Name the proof

Express the next milestone as a falsifiable claim about user value or technical feasibility. If success cannot prove or disprove anything, rewrite it.

Complete when: the milestone can clearly succeed or fail as a learning claim

2. Locate the riskiest assumption

List value, usability, feasibility, viability, and distribution assumptions. Rank by uncertainty multiplied by consequence; select the assumption this release must test.

Complete when: one dominant uncertainty is selected or a reason for multiple is given

3. Trace the shortest proof path

Map the minimum end-to-end user action that generates credible evidence. Preserve a real outcome, even if backstage operations are manual.

Complete when: the slice reaches a meaningful outcome

4. Cut by evidence contribution

For every feature, integration, role, and infrastructure item, ask whether its absence prevents the proof or invalidates the evidence. Delete, defer, simulate, or time-box everything else.

Complete when: every retained item contributes to proof or mandatory trust

5. Protect trust and irreversibility

Restore minimum safety, privacy, data integrity, accessibility, and operational controls. “MVP” is not permission to create avoidable user harm or irreversible data debt.

Complete when: the slice is safe enough to produce valid evidence

6. Write the scope contract

Define in-scope slice, non-goals, manual seams, success/failure thresholds, instrumentation, deadline, and objective expansion triggers. Include a kill or rethink condition.

Complete when: scope and expansion cannot silently grow

Operating rules

  • Optimize for learning per unit of effort, not feature count.
  • A thin slice crosses the whole value path; a layer-only prototype may prove nothing to users.
  • Manual backstage work is acceptable when disclosed and safe.
  • Do not defer the measurement required to know whether the proof worked.
  • Do not call table stakes “nice to have” when their absence invalidates trust or evidence.

Output contract

The final result must contain:

  • Falsifiable proof statement
  • Riskiest-assumption ranking
  • Minimum end-to-end slice
  • Keep/cut/defer/manual decisions with rationale
  • Explicit non-goals
  • Success, expansion, and kill thresholds

Distinguish observed, inferred, assumed, and unverified claims. Link to files, commits, screenshots, commands, logs, or supplied sources when those artifacts support a consequential conclusion. Keep the executive result concise; move detail behind appendices when necessary.

Common pitfalls

  • Tiny full product: shrinking every feature instead of isolating one proof.
  • Infrastructure-first scope that delays contact with the core risk.
  • Vanity metrics disconnected from the claim.
  • Unsafe shortcuts defended as MVP pragmatism.

Verification checklist

  • The requested outcome—not merely the procedure—has been satisfied.
  • Every consequential claim is supported, qualified, or explicitly unverified.
  • The workflow's completion criteria are met in order.
  • Safety boundaries and forbidden side effects were respected.
  • The output follows assets/output-template.md or contains equivalent information.
  • Remaining unknowns have an owner, next action, or stated consequence.

Resources

  • references/playbook.md — deeper domain prompts and evaluation guidance.
  • assets/output-template.md — reusable final-output structure.
  • evals/evals.json — output-quality cases.
  • evals/trigger-evals.json — positive and negative description-trigger cases.

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.