agentleFS
phatnguyen03022001phatnguyen03022001/agent-skills/gap-analysis/SKILL.md

gap-analysis

Skill2 starsChanged 3 months agoApache-2.0View on GitHub

Use when a specification, design, plan, contract, migration, or implementation proposal may be incomplete and missing decisions could block safe execution or verification.

21 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

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

  • 0

    Common

    Most of the 20 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Search the negative space: what must be true for this plan to work that has not been specified?

    line 8
  2. Distinctive:

    Do not fill an architectural gap by guessing.

    line 46
  3. Distinctive:

    Do not manufacture exhaustive checklists for low-risk trivial changes.

    line 54

“Common” is measured: kept by at least 6 in 10 of the 20 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

Gap Analysis

Search the negative space: what must be true for this plan to work that has not been specified?

Build the expected model

Identify actors, inputs, outputs, states, transitions, data ownership, external dependencies, lifecycle stages, invariants, and success criteria. Compare that model with what the artifact actually specifies.

Probe for missing:

  • requirements and acceptance criteria;
  • initial, terminal, invalid, and transitional states;
  • error, timeout, cancellation, retry, and partial-success paths;
  • authorization and trust-boundary decisions;
  • validation and malformed-input behavior;
  • concurrency and idempotency semantics;
  • migrations, compatibility, rollout, rollback, and recovery;
  • cleanup, retention, deletion, and resource ownership;
  • observability, alerts, diagnostics, and operator actions;
  • tests and authoritative verification;
  • dependency assumptions and version constraints;
  • ownership of schemas, configuration, secrets, and generated artifacts;
  • invariants and forbidden changes.

Trace at least one normal path and relevant exceptional paths end-to-end. For stateful changes, ask what happens before, during, and after transition. For integrations, ask what happens when the dependency is slow, unavailable, duplicated, reordered, or returns unexpected data.

Material-design-readiness

For consequential implementation, identify the applicable target product/design authority and surface material missing decisions that could change correctness, security, compatibility, ownership, irreversible behavior, or verification. Resolve those decisions before large or consequential coding. Trivial, mechanical, reversible, or well-specified work should not receive exhaustive documentation ceremony merely because gap analysis exists.

Classify gaps

For every real gap, record:

  • what is missing;
  • why it matters;
  • the decision owner or authority source;
  • whether it blocks execution;
  • what evidence would close it.

Do not fill an architectural gap by guessing. A blocking unknown should remain explicit until resolved or the Architect issues a revised contract.

Prioritize gaps by consequence and decision timing. A missing rollback path before an irreversible migration matters more than an undocumented cosmetic convention. Watch for requirements that exist only implicitly in tests, deployment scripts, schemas, or operational habits; surface them as candidate authority conflicts rather than silently treating implementation drift as intent.

Boundaries

Gap analysis is about completeness, not preference. If a decision is specified but questionable, use design-review. If a specified design needs fault pressure, use adversarial-audit. If a missing fact requires external evidence, use research.

Do not manufacture exhaustive checklists for low-risk trivial changes. Search where missing information could change correctness, safety, compatibility, operations, or verification.

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.