Jarvis Ci Flake Hunter 01

Track and reduce flaky tests with deterministic isolation. Use when handling development work that needs repeatable steps, quality gates, and practical outputs.

Security Scan

Instruction-only, high-level playbook that doesn't request any credentials or install anything; coherent but more general than its CI-specific name suggests.

Purpose & Capability

The skill's name and description claim 'Track and reduce flaky tests with deterministic isolation', but the SKILL.md is a generic development playbook (define success, checkpoints, quality gates, outputs) and contains no CI/test-specific commands, test isolation steps, or references to CI systems. This is likely fine for a human-guided planning assistant but may not meet expectations if you expected automated flake-detection or CI integration.

Instruction Scope

The runtime instructions are high-level guidance only; they do not tell the agent to read arbitrary system files, access environment variables, contact external endpoints, or execute commands. The agent may process any artifacts you explicitly provide (code, logs, screenshots), but the skill itself does not instruct unexpected data collection or transmission.

Install Mechanism

No install spec and no code files — lowest-risk model. Nothing is downloaded or written to disk by the skill manifest.

Credentials

The skill declares no required environment variables, credentials, or config paths. It does not request any secret or system access relative to its stated purpose.

Persistence & Privilege

always is false and there are no indications the skill modifies agent/system configuration or requests persistent presence. Autonomous invocation is enabled by default but not problematic here given the skill's benign footprint.

Assessment

This skill is safe from a permissions perspective — it only contains a generic playbook and asks for inputs you would provide. Be aware it will not itself run CI jobs or perform deterministic isolation; if you need automated flake detection or CI integration, look for a skill that explicitly lists CI commands, required CI tokens, or test-runner integrations. Avoid supplying secrets or CI tokens to the agent unless you trust it and have audited the skill; if you plan to let the agent act on repositories or CI systems, test in a sandboxed environment first and verify any generated commands before running them.

Deliver a concrete, reusable workflow for development tasks.

Inputs

• Desired outcome and deadline
• Constraints (time, tools, risk tolerance)
• Available artifacts (code, docs, screenshots, logs)

Playbook

1. Define success and failure criteria in measurable terms.
2. Build a smallest-valid execution path before scaling effort.
3. Execute in checkpoints; record evidence at each checkpoint.
4. Apply quality gates before finalizing recommendations.
5. Return implementation-ready outputs and next actions.

Quality Gates

• Evidence-backed claims only.
• Explicit assumptions and tradeoffs.
• Reversible changes preferred for risky operations.
• Clear owner + ETA for each next step.
• Include a fallback if the primary plan fails.

Output Format

• Situation summary (5 lines max)
• Top findings (ranked by impact)
• Action plan (today / this week)
• Risks + mitigations
• Exact checklist or commands where useful

Example prompts

• "Use this skill to turn my messy notes into a production-ready plan."
• "Run this skill and give me a risk-first action sequence with acceptance checks."