agentleFS
willianbswillianbs/skills/test-strategy-designer/SKILL.md

test-strategy-designer

Skill0 starsChanged 4 months agoMITView on GitHub

Designs a risk-based test strategy mapped to acceptance criteria without writing test code unless asked. Use after planning and before or during implementation. Emits TEST_STRATEGY. Never chases 100% coverage or tests implementation details by default.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 4 are distinctive, and 2 are 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.

  • 2

    Very specific

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

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

    Prefer SPEC_STATUS with AC IDs + PLAN/TASK_GRAPH.

    line 24
  2. Distinctive:

    Do not write tests unless asked.

    line 46
  3. Very specific:

    Every AC_ID appears in the matrix.

    line 62
  4. Distinctive:

    Never demand 100% line coverage as a goal.

    line 97
  5. Distinctive:

    Never write brittle tests coupled to private implementation unless asked.

    line 98
  6. Distinctive:

    Never skip negative authz cases for auth changes.

    line 99

≈ “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

Purpose

Define what must be tested, at which level, and what blocks CI—so correctness and regressions are intentional, not accidental.

When to Use / When NOT to Use

Use when: feature has ACs; HighRisk change; after delivery-planner; before large implementer work.

Do not use when: user only wants a single unit test written (feature-implementer / tdd); pure docs.

Preconditions

Prefer SPEC_STATUS with AC IDs + PLAN/TASK_GRAPH. If missing, derive provisional ACs and mark them provisional.

Inputs / Outputs

Inputs: AC list, PLAN/TASK_GRAPH, CONTEXT_PACK, risk notes.

Outputs: TEST_STRATEGY

Upstream / Downstream

Upstream: spec-validator, delivery-planner.

Downstream: feature-implementer, quality-gate, defect-analyst (regression).

Core Principles

  1. Risk-based, not coverage-vanity.
  2. Behavior over implementation details.
  3. Pyramid by default; justify every E2E.
  4. Security/perf tests mandatory when risk ≥ High for those domains.
  5. Flakes are defects—define quarantine/ownership.
  6. CI blocking set ≠ nightly set.
  7. Do not write tests unless asked.

Process

  1. Ingest AC IDs and risk (auth, money, data, concurrency).
  2. Build matrix: each AC_ID → unit / integration / e2e / manual / monitoring.
  3. Choose pyramid defaults for the stack found in CONTEXT_PACK; justify deviations.
  4. Mocking strategy — what to fake vs real.
  5. CI blocking vs nightly/scheduled.
  6. Required security tests (authz negative cases) if auth/PII/payments.
  7. Required perf smoke if latency/SLO sensitive.
  8. Flake policy — quarantine, owner, max retries.
  9. Production validation hooks if relevant (feature flag, metric).

Evidence Requirements

Every AC_ID appears in the matrix. “No test needed” requires rationale.

Stop Conditions / Failure Modes

ConditionAction
ACs not testableHand back to spec-validator (Revise)
Only E2E proposed for everythingRevise pyramid

Severity + Confidence

Untested Critical AC → strategy Decision Block until covered in matrix.

Output Contract

## TEST_STRATEGY
Matrix:
  AC-1: unit, integration
Pyramid notes: ...
CI blocking: ...
Nightly: ...
Security/perf required: ...
Flake policy: ...
Decision: Proceed | ProceedWithConditions | Revise | Block

Handoffs

  • feature-implementer — execute strategy
  • quality-gate — verify CI blocking evidence
  • defect-analyst — when failures appear

Never

  • Never demand 100% line coverage as a goal.
  • Never write brittle tests coupled to private implementation unless asked.
  • Never skip negative authz cases for auth changes.

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.