eval-harness
A formal evaluation framework for agent sessions implementing eval-driven development (capability evals, regression evals, graders, pass@k metrics) — the…
Skill by thekiwidev
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/thekiwidev-eval-harness
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: agent-building
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-harness
What's inside
- Skill name
eval-harness- Description
- A formal evaluation framework for agent sessions implementing eval-driven development (capability evals, regression evals, graders, pass@k metrics) — the reference the eval skill operates on.
- License
- NOASSERTION
- Instructions
- 223 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
The license isn't open source, so the instructions aren't shown here. Read SKILL.md on GitHub
More from thekiwidev/ai-system
- 0.593score
create-workflow
thekiwidev
Create or update a repeatable workflow (release, production build, OTA update, deploy, migration, publish to Expo Go, any "every time I do X I must do these…
SkillAFreshness: A★ 2Claude Code +9 - 0.593score
bugfix-workflow
thekiwidev
Workflow for a bug, defect or regression — establish expected behaviour from the authoritative source, reproduce, find the root cause before changing code…
SkillAFreshness: A★ 2Claude Code +9 - 0.593score
feature-workflow
thekiwidev
End-to-end workflow for a NEW product capability or feature — classify, product definition (create-prd) and owner PRD approval, implementation plan and owner…
SkillAFreshness: A★ 2Claude Code +9 - 0.586score
_template-workflow
thekiwidev
One sentence — what this workflow does and the requests that should run it.
SkillAFreshness: A★ 2Claude Code +9 - 0.552score
context-docs
thekiwidev
Regenerate a project's derived, agent-facing context docs (docs/ai/context/*.md — caveman-ultra copies of MEMORY.md, NOTES.md, CHANGELOG.md) from their…
SkillBFreshness: B★ 2Claude Code +9 - 0.552score
create-prd
thekiwidev
Create a detailed Product Requirements Document (PRD) in Markdown from a feature idea — asks 3–10 essential clarifying questions with lettered options, then…
SkillBFreshness: B★ 2Claude Code +9 - 0.549score
build-fix
thekiwidev
Incrementally fix TypeScript/build errors one at a time using the project's real build command; stops when a fix introduces new errors or the same error…
SkillBFreshness: B★ 2Claude Code +9 - 0.549score
code-review
thekiwidev
Security and quality review of uncommitted changes with severity-ranked findings (CRITICAL/HIGH/MEDIUM/LOW), file:line locations and concrete fixes; never…
SkillBFreshness: B★ 2Claude Code +9 - 0.549score
checkpoint
thekiwidev
Create, verify or list workflow checkpoints (named snapshots of test/coverage/build state) so long tasks can be compared against a known-good point.
SkillBFreshness: B★ 2Claude Code +9 - 0.546score
continuous-learning
thekiwidev
Automatically extract reusable patterns from agent sessions and save them as learned skills for future use.
SkillBFreshness: B★ 2Claude Code +9 - 0.540score
_template
thekiwidev
One sentence — what this skill does and WHEN an agent should reach for it.
SkillBFreshness: B★ 2Claude Code +9 - 0.297score
e2e
thekiwidev
No description from the publisher yet.
SkillBFreshness: B★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 21, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- NOASSERTION
- Repository
- github.com/thekiwidev/ai-system
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from thekiwidev
- 0.593score
create-workflow
thekiwidev
Create or update a repeatable workflow (release, production build, OTA update, deploy, migration, publish to Expo Go, any "every time I do X I must do these…
SkillAFreshness: A★ 2Claude Code +9 - 0.593score
bugfix-workflow
thekiwidev
Workflow for a bug, defect or regression — establish expected behaviour from the authoritative source, reproduce, find the root cause before changing code…
SkillAFreshness: A★ 2Claude Code +9 - 0.593score
feature-workflow
thekiwidev
End-to-end workflow for a NEW product capability or feature — classify, product definition (create-prd) and owner PRD approval, implementation plan and owner…
SkillAFreshness: A★ 2Claude Code +9 - 0.586score
_template-workflow
thekiwidev
One sentence — what this workflow does and the requests that should run it.
SkillAFreshness: A★ 2Claude Code +9 - 0.552score
context-docs
thekiwidev
Regenerate a project's derived, agent-facing context docs (docs/ai/context/*.md — caveman-ultra copies of MEMORY.md, NOTES.md, CHANGELOG.md) from their…
SkillBFreshness: B★ 2Claude Code +9 - 0.552score
create-prd
thekiwidev
Create a detailed Product Requirements Document (PRD) in Markdown from a feature idea — asks 3–10 essential clarifying questions with lettered options, then…
SkillBFreshness: B★ 2Claude Code +9
Cite this page
eval-harness skill. Agent Mart. https://agentmart.dev/skills/thekiwidev-eval-harness (last updated 2026-10-24)