eval-driven-testing
Build inspectable evaluations for prompts, skills, agents, and model workflows.
Skill by Simultech369
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/simultech369-eval-driven-testing
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: testing
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/eval-driven-testing
What's inside
- Skill name
eval-driven-testing- Description
- Build inspectable evaluations for prompts, skills, agents, and model workflows. Use before behavior changes or when reliability must be measured across revisions.
- License
- ISC
- Instructions
- 9 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
- Define expected behavior and failure cases before implementation. - Separate deterministic contract tests from model-judged quality evaluations. - Keep single-turn, multi-turn, and tool-using scenarios distinct. - Store fixtures, grader rules, metrics, outputs, and run metadata in inspectable artifacts. - Include adversarial, boundary, stale-context, and no-op cases. - Compare against a baseline and report regressions, not only aggregate scores. - Use human or model rubrics only where deterministic checks cannot express the requirement.
From Simultech369/Dizzy-the-Polymath, ISC.
More from Simultech369/Dizzy-the-Polymath
- 0.544score
lsp-standards
Simultech369
Construct, validate, and interact with on-chain agent identities using ERC-725 and LUKSO Standard Proposals (LSP0 Universal Profile, LSP6 Key Manager, LSP2…
SkillBFreshness: B★ 1Claude Code +9 - 0.542score
rhythm-coaching
Simultech369
Coach drum grooves, fills, counting grids, and practice loops from text or tapped input.
SkillBFreshness: B★ 1Claude Code +9 - 0.492score
context-budgeting
Simultech369
Design a bounded context set for a task while controlling bloat, poisoning, duplication, and lost-in-the-middle risk.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
memory-discipline
Simultech369
Maintain high-signal memory by separating daily logs from durable lessons.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
model-routing
Simultech369
Select and route tasks to the best model by cost, latency, quality, and modality.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
agent-relationships
Simultech369
Coordinate multi-agent collaboration with explicit trust, handoff, and conflict protocols.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
filesystem-workplace
Simultech369
Navigate, inspect, and organize local files and folders safely.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
browser-automation
Simultech369
Automate browser tasks deterministically for testing, scraping, or workflow execution.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
git-skill
Simultech369
Manage code changes, diffs, commits, and branch hygiene safely.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
ci-log-triage
Simultech369
Diagnose CI, workflow, build, test, and release-readiness failures from exact logs and commands.
SkillCFreshness: C★ 1Claude Code +9 - 0.489score
api-contract-design
Simultech369
Design or review stable API contracts, validation, errors, health surfaces, and compatibility behavior.
SkillCFreshness: C★ 1Claude Code +9 - 0.489score
plan-execution-checkpoints
Simultech369
Execute a written implementation plan in bounded verified batches.
SkillCFreshness: C★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jun 11, 2026 (4 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- ISC
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from Simultech369
- 0.544score
lsp-standards
Simultech369
Construct, validate, and interact with on-chain agent identities using ERC-725 and LUKSO Standard Proposals (LSP0 Universal Profile, LSP6 Key Manager, LSP2…
SkillBFreshness: B★ 1Claude Code +9 - 0.542score
rhythm-coaching
Simultech369
Coach drum grooves, fills, counting grids, and practice loops from text or tapped input.
SkillBFreshness: B★ 1Claude Code +9 - 0.492score
agent-relationships
Simultech369
Coordinate multi-agent collaboration with explicit trust, handoff, and conflict protocols.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
memory-discipline
Simultech369
Maintain high-signal memory by separating daily logs from durable lessons.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
git-skill
Simultech369
Manage code changes, diffs, commits, and branch hygiene safely.
SkillCFreshness: C★ 1Claude Code +9 - 0.492score
browser-automation
Simultech369
Automate browser tasks deterministically for testing, scraping, or workflow execution.
SkillCFreshness: C★ 1Claude Code +9
Cite this page
eval-driven-testing skill. Agent Mart. https://agentmart.dev/skills/simultech369-eval-driven-testing (last updated 2026-10-24)